00:12:21  * TooTallNatequit (Ping timeout: 244 seconds)
00:15:47  * TooTallNatejoined
00:23:33  * EhevuTovquit (Quit: This computer has gone to sleep)
00:26:30  <isaacs>wow, i'm seeing some really baffling behavior from ab
00:26:39  <isaacs>i thought it was my program, but now i'm not so sure.
00:26:42  <isaacs>unless nginx has the same issue.
00:28:44  <AvianFlu>I've always been extremely distrustful of ab, but I have no proof to such
00:29:08  * pooyaquit (Quit: pooya)
00:30:36  <isaacs>ok, i just saw it go from 10,000 q/s to 300q/s against nginx
00:30:39  <isaacs>that's not reasonable.
00:32:29  <isaacs>it's like it occasionally just gets "stuck"
00:39:38  <isaacs>hm... seems like doing sendfile in an http server is just not possible now...
00:39:44  <isaacs>sine we don't have a fd on the socket
00:43:47  * mikealquit (Quit: Leaving.)
00:51:21  * mcavagejoined
00:52:11  * piscisaureus_joined
00:52:26  * stephankquit (*.net *.split)
00:55:36  * perezdquit (Quit: perezd)
00:58:58  * stephankjoined
01:03:07  * mikealjoined
01:06:41  * joshthecoderquit (Quit: Leaving...)
01:09:59  * piscisaureus_quit (Ping timeout: 272 seconds)
01:10:45  * c4miloquit (Remote host closed the connection)
01:18:30  * avalanche123quit (Quit: Textual IRC Client: http://www.textualapp.com/)
01:35:55  * beachdog_quit (Remote host closed the connection)
01:37:39  * piscisaureus_joined
01:39:17  * mcavagequit (Remote host closed the connection)
02:03:14  * chobi_e_changed nick to chobi_e
02:03:24  * c4milojoined
02:05:03  * piscisaureus_quit (Ping timeout: 240 seconds)
02:08:52  * mcavagejoined
02:12:48  * lohkeyquit (Quit: lohkey)
02:24:31  * mcavagequit (Remote host closed the connection)
02:25:54  * mikealquit (Quit: Leaving.)
02:27:26  * TooTallNatequit (Quit: Computer has gone to sleep.)
02:43:44  * bradleymeckquit (Quit: bradleymeck)
02:46:29  * blackorzar_quit (Ping timeout: 276 seconds)
03:16:17  * mmaleckiquit (Ping timeout: 244 seconds)
03:17:17  * mcavagejoined
03:27:11  * yanushquit (Ping timeout: 245 seconds)
03:29:35  * brsonquit (Quit: leaving)
03:38:15  * c4miloquit (Remote host closed the connection)
03:44:02  * dshaw_quit (Quit: Leaving.)
03:47:57  * mcavagequit (Remote host closed the connection)
03:56:54  * brsonjoined
03:57:26  * brsonquit (Client Quit)
03:57:33  * brsonjoined
03:59:46  * chobi_echanged nick to chobi_e_
04:05:23  * pooyajoined
04:15:03  * pooyaquit (Quit: pooya)
04:19:18  * mikealjoined
04:34:25  * pooyajoined
05:00:29  * pooyaquit (Quit: pooya)
05:08:39  * mikealquit (Quit: Leaving.)
05:26:38  * theColejoined
05:31:30  * avalanche123joined
05:49:29  * stagasjoined
05:59:32  * blackorzarjoined
05:59:44  * joshthecoderjoined
06:17:40  * avalanche123quit (Quit: Computer has gone to sleep.)
06:30:58  * joshthecoderquit (Quit: Leaving...)
06:31:15  * mmaleckijoined
06:36:45  * AvianFluquit (Quit: AvianFlu)
06:58:47  * mikealjoined
07:05:38  * stagasquit (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.0.17/2009122204])
07:38:49  * stagasjoined
07:40:33  * brsonquit (Ping timeout: 252 seconds)
07:40:35  <devaholic>ab is unreliable on some systems =/
07:40:45  * rendarjoined
07:42:04  * blackorzarquit (Ping timeout: 248 seconds)
07:53:17  * stagasquit (Ping timeout: 276 seconds)
08:28:22  * `3rdEdenjoined
08:33:38  * dshaw_joined
08:40:28  * stagasjoined
09:26:12  * stagasquit (Quit: Bye)
09:50:16  * hzjoined
09:53:45  * stagasjoined
10:00:46  * mikeal1joined
10:09:58  * mikealquit (*.net *.split)
10:09:58  * Benviequit (*.net *.split)
10:10:19  * Benviejoined
10:19:56  * dshaw_quit (Quit: Leaving.)
10:26:21  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
10:30:09  <indutny>hey people
10:30:15  <indutny>anyone up for testing vock conference calls?
10:30:24  <indutny>fixed many issues, improved many stuff
10:30:28  <indutny>like echo cancellation
10:43:52  * pooyajoined
10:43:57  * pooyaquit (Client Quit)
11:05:05  * `3rdEdenjoined
11:05:07  * AvianFlujoined
11:35:43  * devaholicquit (Ping timeout: 246 seconds)
11:37:24  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
12:00:20  * AndreasMadsenjoined
12:11:11  * hzquit (Ping timeout: 272 seconds)
12:31:16  * stagas_joined
12:32:44  * stagasquit (Ping timeout: 248 seconds)
12:32:56  * stagas_changed nick to stagas
12:33:06  * theColequit (Quit: theCole)
12:56:07  * hzjoined
13:10:45  * AndreasMadsenquit (Remote host closed the connection)
13:42:05  * hzquit
13:59:51  * AvianFlu_joined
14:03:08  * AvianFluquit (Ping timeout: 276 seconds)
14:03:09  * AvianFlu_changed nick to AvianFlu
14:25:58  * beachdogjoined
14:37:56  * AvianFluquit (Quit: AvianFlu)
14:53:15  * blackorzarjoined
15:11:31  * AvianFlujoined
15:12:51  * bnoordhuisjoined
15:23:50  * AvianFluquit (Quit: AvianFlu)
15:30:31  * stagasquit (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.0.17/2009122204])
15:45:16  * AvianFlujoined
15:59:27  * toothrchanged nick to toothrot
16:13:15  <CIA-134>node: isaacs v0.6 * r7e6dea4 / Makefile : build: Regenerate docs for tarball and releases (+7 more commits...) - http://git.io/FgucAw
16:17:52  <tjfontaine>bnoordhuis: should setImmediate also adhere to ref/unref? (I think not, but it makes for a weird api)
16:20:00  <bnoordhuis>tjfontaine: you mean should a setImmediate handle be unrefable?
16:20:07  <tjfontaine>yes
16:20:19  <bnoordhuis>no. setTimeout and setInterval handles aren't either
16:20:33  <bnoordhuis>(i think)
16:20:39  <tjfontaine>no, I made them unrefable
16:20:55  <tjfontaine>with the sockets unref patch
16:21:19  <bnoordhuis>ah
16:21:20  <bnoordhuis>why?
16:21:28  <bnoordhuis>i see bert merged it
16:21:33  <tjfontaine>because it's a handle that can be unref'd?
16:21:46  <tjfontaine>if you have some gc like setInterval for instance
16:21:54  <bnoordhuis>okay
16:22:02  <bnoordhuis>i'm not opposed, just curious about the motivation
16:22:16  <isaacs>actually setTimeout is a good candidate for unrefableness
16:22:25  <bnoordhuis>it already is
16:22:32  <bnoordhuis>i.e. not a candidate anymore
16:22:39  <isaacs>heh, right
16:22:43  <isaacs>but i mean, a very good use case
16:23:04  <isaacs>setTimeout(makeSureItFinishedInTenSeconds, 1000).unref()
16:23:16  <isaacs>and just clear the timeout when you finsh
16:23:16  <tjfontaine>so the question is how weird/improper is it if setImmediate can't be unref'd?
16:23:38  <bnoordhuis>tjfontaine: it's not consistent but making it unrefable seems kind of pointless
16:23:46  <tjfontaine>I agree
16:23:50  <bnoordhuis>let's just wait for the bug reports to come in :)
16:23:54  <tjfontaine>hehe ok
16:25:54  <bnoordhuis>isaacs: i don't understand npm
16:26:02  <bnoordhuis>i have node and npm installed in $HOME/opt/node
16:26:26  <bnoordhuis>when i `npm install -g sm` (don't ask) it installs sm to $HOME/opt/node/lib/node_modules/sm and the deps to $HOME/.npm
16:26:54  <bnoordhuis>but the odd thing is, on my laptop it seems to install the deps to $HOME/local/lib (iirc, it was $HOME/local/something)
16:27:03  <bnoordhuis>same node version btw
16:33:21  * AvianFluquit (Read error: Connection reset by peer)
16:33:42  * AvianFlujoined
16:36:02  * dshaw_joined
16:38:09  * mmaleckiquit (Ping timeout: 245 seconds)
16:42:11  <isaacs>bnoordhuis: ~/.npm is a cache folder -- it's not "installing" things there, just unpacking them temporarily to work with
16:42:38  <isaacs>and mostly, its behavior in that area is a vestigial throwback to the days when npm used system command-line tar utilities
16:42:47  <bnoordhuis>isaacs: ah, right - that makes sense
16:43:02  <isaacs>the deps would be in ~/opt/node/lib/node_modules/sm/node_modules/
16:43:43  <bnoordhuis>okay, cool. where does $HOME/local come into play?
16:44:06  <bnoordhuis>"i don't know" is a perfectly valid answer, it could be an install script
16:54:32  <isaacs>yeah, i don't know about home/local
16:54:47  <isaacs>it defaults the -g root to {node's execPath}/../lib/node_modules/
16:55:14  <isaacs>but it is also configurable
16:55:22  <isaacs>npm config ls <-- will show you what's set where
17:02:15  * piscisaureus_joined
17:02:21  <piscisaureus_>hey bnoordhuis
17:02:28  <bnoordhuis>ho piscisaureus_
17:02:39  <piscisaureus_>can you update heapdump's package.json to use build/Release/heapdump.node for main ?
17:03:42  <bnoordhuis>piscisaureus_: i take pull requests :)
17:03:52  <bnoordhuis>(i kid, i kid - i'll fix it)
17:04:48  <piscisaureus_>bnoordhuis: I did it locally, but I dont like sending PRs for one-liner fixes
17:06:18  * theColejoined
17:08:15  <bnoordhuis>piscisaureus_: done, npm install [email protected]
17:08:22  <piscisaureus_>bnoordhuis: thanks!
17:11:28  * AvianFluquit (Quit: AvianFlu)
17:15:22  <isaacs>bnoordhuis, piscisaureus_: how would you feel about blessing the "finish" method on http responses?
17:15:49  <isaacs>we already rely on it internally
17:16:03  <bnoordhuis>isaacs: for what?
17:16:11  <bnoordhuis>i mean, why bless it?
17:16:37  * mcavagejoined
17:17:17  * theColequit (Quit: theCole)
17:18:54  <bnoordhuis>piscisaureus_: why does c9 want to 'Update your user profile.' (on github)?
17:19:06  <piscisaureus_>bnoordhuis: does it?
17:19:14  <bnoordhuis>when you want to sign up through github
17:19:24  <piscisaureus_>bnoordhuis: ooh - I think it adds ssh keys
17:19:36  <piscisaureus_>bnoordhuis: so cloud9 can list your projects
17:19:40  <bnoordhuis>ah
17:19:55  <bnoordhuis>i guess i'll try a regular signup first :)
17:26:49  <bnoordhuis>ah, good ol' RangeError: Maximum call stack size exceeded
17:26:50  <isaacs>bnoordhuis: because it's useful to know when the response has been ended.
17:27:07  <bnoordhuis>isaacs: do you mean the 'finish' event?
17:27:13  * TooTallNatejoined
17:27:16  <isaacs>bnoordhuis: oh, righ, event, not method
17:27:25  <isaacs>there is no finish() method, i don't think :)
17:27:44  <isaacs>thought, maybe it should be called "end" anyway
17:27:57  <bnoordhuis>isaacs: i'm fine with that and yes, 'end' is a better name
17:28:39  <bnoordhuis>makes for parity with the 'end' response event
17:28:43  <isaacs>the question is: is it more confusing to have different names for the same concept (being done), or is it more confusing to use the same thing to mean two different things (no more data coming out vs no more data going in)
17:28:50  <isaacs>yeah, i kinda lean that way, as well
17:29:09  <isaacs>and then we can make this userland module obsolete: http://npm.im/response-end
17:30:06  <bnoordhuis>sounds good to me
17:32:49  <isaacs>does 'finish' get emitted right away when you call end, or only once the write is flushed?
17:33:17  <indutny>piscisaureus_: hey bertje
17:33:23  <indutny>piscisaureus_: are you in osx atm?
17:33:31  <piscisaureus_>indutny: nope
21:34:33  * piscisaureus_joined
21:34:55  <piscisaureus_>hmm, joyent having connectivity issues?
21:38:53  <bnoordhuis>indutny: did you rebase #499?
21:41:51  <bnoordhuis>ah, don't worry about it - resolved the conflict
21:45:28  <CIA-134>libuv: Fedor Indutny master * r23dc564 / (include/uv-private/uv-unix.h src/unix/fs.c): darwin: emulate fdatasync() with fcntl(F_FULLFSYNC) - http://git.io/M5jRiw
21:47:05  * travis-cijoined
21:47:06  <travis-ci>[travis-ci] joyent/libuv#545 (master - 23dc564 : Fedor Indutny): The build passed.
21:47:06  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/837edf4c0f60...23dc564f3baf
21:47:06  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2096902
21:47:06  * travis-cipart
21:53:45  <mscdex>bnoordhuis: thoughts on my PR?
21:53:53  <bnoordhuis>mscdex: which one?
21:54:17  <mscdex>the one about making more parser errors emit on http req object instead of only at the socket level
21:54:45  <bnoordhuis>iirc it lgtm'd last time i looked at it
21:55:21  <bnoordhuis>i'm a bit behind on reviewing patches though
21:55:49  <mscdex>ok
21:56:41  * dshaw_joined
21:59:25  <bnoordhuis>tjfontaine: your PR breaks test/simple/test-process-active-wraps.js for reasons i don't fully comprehend
21:59:47  <bnoordhuis>that test is somewhat outdated after the nextTick refactor though
21:59:56  <tjfontaine>hmm I shall investigate
22:01:24  <tjfontaine>oh
22:01:25  <bnoordhuis>ah, i think it's the var immediateTimer = new Timer();
22:01:32  <tjfontaine>yup
22:02:00  <tjfontaine>I could lazily create the timer, which fixes that test
22:03:50  <tjfontaine>or I could set the assert to check for 1 :)
22:04:06  <bnoordhuis>there's a pummel test that checks for it as well
22:04:16  <bnoordhuis>no worries, i'll fix it up in lib/timers.js
22:04:32  <tjfontaine>ok
22:04:44  <bnoordhuis>immediateLazyInit() is kind of an odd name for a function...
22:04:52  <tjfontaine>heh
22:08:22  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
22:08:58  <bnoordhuis>tjfontaine: https://github.com/bnoordhuis/node/commit/382f22f <- lgty?
22:09:37  <tjfontaine>bnoordhuis: yup
22:10:12  <CIA-134>node: Kyle Robinson Young v0.8 * r4ef808e / doc/api/child_process.markdown : doc: remove unused util from child_process - http://git.io/k2FZjg
22:10:40  <CIA-134>node: Timothy J Fontaine master * r382f22f / (5 files in 5 dirs): timers: implement setImmediate - http://git.io/YcTVaA
22:11:15  <tjfontaine>there is a small bug we will want to fix in lib/timers.js wrt unref+domains
22:11:53  * AvianFlujoined
22:12:51  <tjfontaine>https://github.com/bnoordhuis/node/blob/master/lib/timers.js#L258 we should probably this._handle.domain = this.domain
22:13:17  <bnoordhuis>tjfontaine: we take pull requests :)
22:13:21  <tjfontaine>hehe
22:14:11  <bnoordhuis>do we actually set _handle.domain anywhere?
22:14:23  <tjfontaine>best analog is setInterval
22:16:09  <bnoordhuis>tjfontaine: yes, but that creates a new timer each time
22:16:40  <tjfontaine>bnoordhuis: in the case of .uref we're creating a new timer but losing the domain
22:17:08  <bnoordhuis>ah right
22:17:32  <bnoordhuis>a pull request with tests would be appreciated :)
22:17:45  <tjfontaine>hehe, ya this will mean I'll have to "learn" domains :)
22:17:57  <tjfontaine>but ya I will do this
22:18:43  * piscisaureus_joined
22:32:04  * Ariajoined
22:36:28  <tjfontaine>bnoordhuis: https://github.com/joyent/node/pull/3858/files :)
22:47:30  * mmaleckijoined
23:05:57  <CIA-134>libuv: Ben Noordhuis master * rac0d468 / src/unix/process.c : unix: work around darwin bug, don't poll() on pipe - http://git.io/F_t24g
23:09:04  * rendarquit
23:10:28  * AvianFlu_joined
23:11:00  * AvianFluquit (Disconnected by services)
23:11:01  * AvianFlu_changed nick to AvianFlu
23:12:55  * mmaleckiquit (Ping timeout: 268 seconds)
23:14:59  * travis-cijoined
23:14:59  <travis-ci>[travis-ci] joyent/libuv#546 (master - ac0d468 : Ben Noordhuis): The build passed.
23:14:59  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/23dc564f3baf...ac0d46851ae0
23:14:59  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2097304
23:14:59  * travis-cipart
23:18:12  * mmaleckijoined
23:18:21  <CIA-134>node: Timothy J Fontaine master * r320ae69 / (lib/timers.js test/simple/test-domain-timers.js): timers: make unref'd timeouts execute in a domain - http://git.io/E-4syg
23:19:49  * AvianFlu_joined
23:20:03  * blackorzar__quit (Ping timeout: 240 seconds)
23:20:40  * AvianFluquit (Disconnected by services)
23:20:41  * AvianFlu_changed nick to AvianFlu
23:25:03  * brsonquit (Quit: leaving)
23:28:38  * dshaw_quit (Quit: Leaving.)
23:29:25  * AvianFlu_joined
23:30:22  * AvianFluquit (Disconnected by services)
23:30:23  * AvianFlu_changed nick to AvianFlu
23:33:16  * mmaleckiquit (Ping timeout: 252 seconds)
23:38:16  * joshthecoderjoined
23:42:44  * blackorzarjoined