00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:07  * ircretaryjoined
00:04:38  <mordy__>hrm, another question -- how do i determine if i'm using 0.8 or 0.10?
00:04:52  * groundwaterquit (Quit: groundwater)
00:04:55  <mordy__>the 0.8 has a version macro, i believe that was removed though in 0.10
00:05:16  <tjfontaine>uv_version
00:05:19  <tjfontaine>is a function
00:05:27  <mordy__>but, at compile-time
00:06:24  <tjfontaine>I guess check if the version macro exists, if it does you're on 0.8 :)
00:06:43  <mordy__>:)
00:06:58  <mordy__>is there a plan to re-add some kind of version macro in 0.12?
00:07:13  <tjfontaine>I forget why that was even removed
00:24:25  * defunctzombiechanged nick to defunctzombie_zz
00:31:05  * dapquit (Quit: Leaving.)
00:47:18  * piscisaureus_joined
00:48:12  * piscisaureus__joined
00:52:07  * piscisaureus_quit (Ping timeout: 264 seconds)
00:58:05  * indexzerojoined
00:58:06  * qardquit (Quit: Leaving.)
01:00:20  * piscisaureus__quit (Ping timeout: 256 seconds)
01:00:37  * kazuponjoined
01:01:18  * mikealquit (Quit: Leaving.)
01:02:40  * kazupon_joined
01:02:41  * kazuponquit (Read error: Connection reset by peer)
01:04:25  * felixgequit (Quit: felixge)
01:06:19  * amartensquit (Quit: Leaving.)
01:07:18  * amartensjoined
01:11:22  * amartensquit (Client Quit)
01:17:55  * kazupon_quit (Remote host closed the connection)
01:19:46  * abraxasjoined
01:21:00  * kazuponjoined
01:25:56  * kazuponquit (Remote host closed the connection)
01:31:48  * kazuponjoined
01:32:20  * qardjoined
01:42:10  * abraxasquit (Remote host closed the connection)
01:42:26  * abraxasjoined
01:58:23  * paulfryzeljoined
01:58:40  * brsonquit (Quit: leaving)
01:59:44  * c4milojoined
02:06:14  * TooTallNatejoined
02:14:30  * indexzeroquit (Quit: indexzero)
02:15:10  * TooTallNatequit (Read error: Connection reset by peer)
02:20:43  * TooTallNatejoined
02:33:44  * paulfryzelquit (Remote host closed the connection)
02:51:11  * c4miloquit (Remote host closed the connection)
02:51:39  * c4milojoined
02:56:14  * c4miloquit (Ping timeout: 256 seconds)
03:01:27  * indexzerojoined
03:02:54  * inolenquit (Quit: Leaving.)
03:04:53  * AvianFlujoined
03:11:49  * brsonjoined
03:16:49  * julianduquequit (Ping timeout: 246 seconds)
03:25:18  * bill1024joined
03:29:56  * bill1024quit (Ping timeout: 256 seconds)
03:30:35  * bill1024joined
03:30:36  * bill1024quit (Client Quit)
03:49:04  * indexzeroquit (Quit: indexzero)
04:10:01  * indexzerojoined
04:40:33  * indexzeroquit (Quit: indexzero)
04:48:21  * brsonquit (Quit: leaving)
04:48:37  * brsonjoined
04:54:17  * TooTallNatequit (Quit: Computer has gone to sleep.)
04:56:21  * kazuponquit (Remote host closed the connection)
04:57:10  * TooTallNatejoined
05:00:53  * paddybyersjoined
05:07:51  * mikealjoined
05:08:13  * mikealquit (Client Quit)
05:09:00  * AvianFluquit (Remote host closed the connection)
05:11:02  * kazuponjoined
05:13:13  * mikealjoined
05:14:04  * mikealquit (Client Quit)
05:21:42  * bnoordhuisjoined
05:23:33  * defunctzombie_zzchanged nick to defunctzombie
05:24:54  * zpmorganjoined
05:27:48  * brsonquit (Quit: leaving)
05:34:22  * paddybyersquit (Ping timeout: 246 seconds)
06:16:47  * qardquit (Quit: Leaving.)
06:25:39  * felixgejoined
06:41:15  * Damn3dquit (Ping timeout: 245 seconds)
06:46:33  * Damn3djoined
06:47:07  * qardjoined
06:52:48  * bnoordhuisquit (Ping timeout: 268 seconds)
06:54:30  * qardquit (Ping timeout: 264 seconds)
06:55:15  * mikealjoined
07:14:03  * bnoordhuisjoined
07:22:14  * rendarjoined
07:23:35  * defunctzombiechanged nick to defunctzombie_zz
07:24:42  * inolenjoined
07:27:18  * bajtosjoined
07:27:51  * TooTallNatequit (Quit: Computer has gone to sleep.)
07:28:41  * csaohjoined
07:34:17  * bajtosquit (Quit: bajtos)
07:46:00  * hzjoined
07:51:58  * qardjoined
07:53:53  * inolenquit (Quit: Leaving.)
07:55:39  <indutny>hohoho
07:56:15  * qardquit (Ping timeout: 245 seconds)
08:01:10  * defunctzombie_zzchanged nick to defunctzombie
08:05:41  * inolenjoined
08:26:04  <bnoordhuis>sup fedor?
08:41:08  * dominictarrjoined
08:52:27  * qardjoined
08:54:47  * mikealquit (Quit: Leaving.)
08:57:09  * qardquit (Ping timeout: 264 seconds)
08:57:52  * hzquit
09:01:44  * dominictarrquit (Quit: dominictarr)
09:03:01  * mburnsjoined
09:06:28  * defunctzombiechanged nick to defunctzombie_zz
09:08:40  * paddybyersjoined
09:25:00  * mikealjoined
09:30:30  * kazuponquit (Remote host closed the connection)
09:47:07  <indutny>hey bnoordhuis :)
09:47:13  <indutny>I'm rewriting all our code base
09:47:19  <indutny>sadly not in rust, but in linted C++/C
09:58:04  * raljoined
10:04:47  <bnoordhuis>indutny: wait, before you do that lets agree on the cpplint switches first
10:04:55  <indutny>hey man
10:04:58  <indutny>where have you been :)
10:05:11  <indutny>I've already processed half of it
10:05:13  <bnoordhuis>here, in front of my keyboards, fixing all the bugs
10:05:26  <indutny>there was not much fixes and I disabled only one check
10:05:28  <indutny>short/long types
10:05:34  <indutny>because node_crypto.cc uses them
10:05:49  <bnoordhuis>what check is that? type narrowing?
10:06:09  <indutny># 'Use int16/int64/etc, rather than the C type %s' % match.group(1))
10:06:35  <bnoordhuis>so, what's the c type? :)
10:07:00  <bnoordhuis>oh, it suggests you should use int64_t rather than long
10:07:05  <indutny>I can't
10:07:11  <indutny>because openssl has long in signatures
10:07:26  <bnoordhuis>right, i can live with that
10:07:31  <indutny>indeed
10:07:38  <indutny>all other stuff seems to be pretty harmless
10:07:43  <indutny>and changes to code are mostly minimal
10:07:47  <bnoordhuis>it's not bad advice though
10:08:11  <indutny>yeah, I know
10:08:17  <bnoordhuis>maybe we can use int32_t in the future
10:08:47  <bnoordhuis>all our inputs fit in 31 bits and it'll get widened to long if long == 64 bits anyway
10:08:54  <bnoordhuis>anyway, not that important
10:11:49  <indutny>yep
10:13:16  * abraxasquit (Remote host closed the connection)
10:20:01  * Raltjoined
10:36:06  <indutny>bnoordhuis: https://github.com/joyent/node/pull/5957
10:44:55  * csaohquit (Quit: csaoh)
10:52:53  * qardjoined
10:53:00  * hzjoined
10:57:37  * qardquit (Ping timeout: 268 seconds)
10:57:56  * csaohjoined
11:32:58  <bnoordhuis>indutny: i see your pull request and raise you an issue: https://github.com/joyent/node/issues/5959
11:33:28  <indutny>haha
11:37:44  * bajtosjoined
11:42:05  <bnoordhuis>off to city hall, biab
11:43:52  * hzquit (Ping timeout: 268 seconds)
11:47:19  * bnoordhuisquit (Ping timeout: 264 seconds)
11:53:16  * qardjoined
11:54:25  * bajtosquit (Quit: bajtos)
11:57:40  * qardquit (Ping timeout: 256 seconds)
12:01:28  * dominictarrjoined
12:10:52  * dominictarrquit (Quit: dominictarr)
12:11:34  * hzjoined
12:51:52  * hzquit (Ping timeout: 246 seconds)
12:52:33  * bnoordhuisjoined
12:53:37  * qardjoined
12:55:48  <roxlu>indutny: I found a solution for that silly bug
12:56:48  <roxlu>give me a poke if you want to know
12:57:10  * bnoordhuisquit (Ping timeout: 256 seconds)
12:57:55  * qardquit (Ping timeout: 245 seconds)
12:59:23  * jmar777joined
13:13:40  <indutny>roxlu: I want to know :)
13:14:02  <indutny>roxlu: please pm me, I'll read it later
13:14:38  <roxlu>ok
13:27:17  * dominictarrjoined
13:28:00  * bnoordhuisjoined
13:48:36  * pachetjoined
13:48:43  * pachetquit (Client Quit)
13:49:10  * pachetjoined
13:49:10  * pachetquit (Changing host)
13:49:10  * pachetjoined
13:54:00  * qardjoined
13:58:43  * qardquit (Ping timeout: 264 seconds)
14:06:05  * hzjoined
14:18:44  <bnoordhuis>amusing. isaacs has more joyent/node commits but i have more lines touched
14:19:12  <bnoordhuis>1546 vs 1184 and 7198 vs 8033
14:19:38  <bnoordhuis>what's also amusing is that fedor has only 280 commits but 7411 lines touched
14:19:51  <bnoordhuis>good job, fedor :)
14:24:05  <bnoordhuis>and as for libuv
14:24:05  <bnoordhuis>21110 Bert Belder
14:24:05  <bnoordhuis>19134 Ben Noordhuis
14:24:05  <bnoordhuis>7768 Igor Zinkovsky
14:24:05  <bnoordhuis>4904 Ryan Dahl
14:24:08  <bnoordhuis>1508 Fedor Indutny
14:24:17  * AvianFlujoined
14:24:45  * AvianFluquit (Remote host closed the connection)
14:25:13  * AvianFlujoined
14:26:04  <bnoordhuis>'lines touched' is a wrong choice of words btw, 'lines of code in HEAD that git blame attributes to you'
14:37:12  * julianduquejoined
14:43:16  * c4milojoined
14:44:44  * austojoined
14:50:14  * paulfryzeljoined
14:54:22  * qardjoined
14:59:13  * Cheeryjoined
14:59:21  * qardquit (Ping timeout: 268 seconds)
15:03:09  * mikealquit (Quit: Leaving.)
15:04:18  * bradleymeckjoined
15:07:24  * indexzerojoined
15:17:55  * bnoordhuisquit (Ping timeout: 245 seconds)
15:23:06  * c4miloquit (Remote host closed the connection)
15:23:32  * c4milojoined
15:25:22  * c4miloquit (Read error: No route to host)
15:25:38  * c4milojoined
15:32:17  * mikealjoined
15:43:15  <isaacs>indutny: thanks for taking this on!
15:44:17  * bnoordhuisjoined
15:50:21  * piscisaureus__joined
15:51:05  * paulfryz_joined
15:51:24  <piscisaureus__>bnoordhuis: so yay I'm still on #1 on *some* list :-p
15:51:30  <piscisaureus__>not for long I'm afraid
15:52:58  * Raltquit (Quit: Bye)
15:53:13  * Raltjoined
15:55:07  * paulfryzelquit (Ping timeout: 264 seconds)
16:01:53  * julianduquequit (Quit: leaving)
16:02:28  * julianduquejoined
16:03:47  * pachet_joined
16:05:26  <bnoordhuis>piscisaureus__: i'm catching up :)
16:06:02  <bnoordhuis>indutny: did you fix the cpplint things?
16:06:44  * pachetquit (Ping timeout: 256 seconds)
16:07:40  * indexzeroquit (Quit: indexzero)
16:08:23  * julianduquequit (Quit: leaving)
16:11:21  * pachet_changed nick to pachet
16:11:25  * pachetquit (Changing host)
16:11:25  * pachetjoined
16:13:51  * julianduquejoined
16:19:20  <trevnorris>morning
16:21:01  * qardjoined
16:21:17  * trevnorrisquit (Quit: IRCRelay - http://ircrelay.com)
16:24:21  * dapjoined
16:25:47  <MI6>joyent/node: Trevor Norris master * 6a5a7b0 : node: add inTick and lastThrew to infoBox - http://git.io/-xislQ
16:28:35  * AndreasMadsenjoined
16:28:38  * mikealquit (Quit: Leaving.)
16:39:02  <MI6>nodejs-master: #338 UNSTABLE smartos-x64 (10/617) smartos-ia32 (2/617) osx-ia32 (1/617) http://jenkins.nodejs.org/job/nodejs-master/338/
16:39:08  * 13WAAVRJVjoined
16:39:08  * 65MAARQRQjoined
16:39:38  * 13WAAVRJVquit (Client Quit)
16:40:29  * mcavagejoined
16:41:19  <piscisaureus__>bnoordhuis: hey, do you know from the top of your head if indexed slots in objects (not arrays) are also faster than named properties?
16:42:13  * 65MAARQRQchanged nick to trevnorris
16:43:13  <MI6>nodejs-master-windows: #140 FAILURE windows-x64 (19/617) http://jenkins.nodejs.org/job/nodejs-master-windows/140/
16:43:38  <bnoordhuis>piscisaureus__: they are
16:43:45  <bnoordhuis>by a considerable margin too
16:43:50  <piscisaureus__>bnoordhuis: cool
16:44:03  <trevnorris>bnoordhuis: thanks for the reviews.
16:44:13  <bnoordhuis>no problem, trevor
16:44:14  * csaohquit (Quit: csaoh)
16:44:19  <tjfontaine>jenkins, you're, fucking, amazing
16:44:20  <tjfontaine>http://jenkins.nodejs.org/job/nodejs-master-windows/140/DESTCPU=ia32,label=windows/console
16:44:40  * trevnorriscelebrates first c++ patch that didn't need changes
16:45:07  <trevnorris>tjfontaine: process leaked file descriptors?
16:45:26  <tjfontaine>who the fuck knows
16:45:57  <tjfontaine>probably their git-client did something stupid (again)
16:50:57  * TooTallNatejoined
16:52:56  <trevnorris>bnoordhuis: for the array types, think it'd make more sense to do Int8, Uint16, Uint8Clamped instead of Byte, UnsignedShort, Pixel, etc?
16:54:14  <bnoordhuis>trevnorris: yeah, that sounds sensible, typed array-ish
16:55:00  <bnoordhuis>trevnorris: btw, i'd rename that ExtSize() function to something that's a little more descriptive
16:55:19  <bnoordhuis>(seeing that you're removing the ExtType typedef, right?)
16:55:29  * julianduquequit (Quit: leaving)
16:55:37  <trevnorris>bnoordhuis: sounds good, and yeah. it's gone. :)
16:56:45  * c4miloquit (Remote host closed the connection)
16:57:11  * c4milojoined
17:02:16  * c4miloquit (Ping timeout: 264 seconds)
17:05:06  * brianc_joined
17:05:36  * jmar777quit (Read error: Connection reset by peer)
17:05:49  <brianc_>Hey this is more of a node question than a libuv question but....node v0.11.x has removed the NODE_PSYMBOL macro. Any #protips on what to use instead?
17:06:03  * jmar777joined
17:08:59  <bnoordhuis>brianc_: depends on whether you truly need a persistent symbol
17:09:06  <brianc_>don't think I do
17:09:15  <brianc_>it's just strings I'm using as event names in C++ event emitters
17:09:37  <bnoordhuis>then i'd just replace it with String::New() in the places where you use the symbol
17:09:44  <brianc_>I guess I'll just use String::New("eventName") or whatev
17:09:50  <brianc_>sweet - thanks bnoordhuis. :)
17:09:53  <bnoordhuis>np :)
17:10:38  * piscisaureus__quit (Quit: ~ Trillian Astra - www.trillian.im ~)
17:10:42  <brianc_>oh one other thing.... FunctionTemplate::New is deprecated. What's the standard way to do that now?
17:11:02  <brianc_>If it's easier you can just point me to the v8 documentation on it. Last time I went snooping for v8 documentation it was scarce, but that was quite a while ago
17:13:26  <brianc_>actually forget it, I'll quit being lazy & go figure it out myself.
17:13:28  <brianc_><3 thanks again
17:13:43  * AvianFluquit (Remote host closed the connection)
17:14:10  <bnoordhuis>brianc_: FunctionTemplate::New() hasn't been deprecated but it's signature changed a bit
17:14:39  * mikealjoined
17:14:39  <bnoordhuis>you're probably calling the version that uses InvocationCallback rather than FunctionCallback
17:15:03  <bnoordhuis>i guess that's not clarifying things at all, is it?
17:15:14  <brianc_>haha
17:15:18  <bnoordhuis>if you look at the definition in v8.h it'll make sense :)
17:15:22  <brianc_>k sweet thanks. :)
17:27:19  * defunctzombie_zzchanged nick to defunctzombie
17:28:42  * defunctzombiechanged nick to defunctzombie_zz
17:31:31  * c4milojoined
17:44:03  * AvianFlujoined
17:46:06  * c4miloquit (Remote host closed the connection)
17:46:32  * c4milojoined
17:49:19  * hueniversequit (Quit: Leaving.)
17:49:29  * c4milo_joined
17:49:43  * c4miloquit (Read error: No route to host)
17:50:14  * Benvie_joined
17:52:19  * Benviequit (Ping timeout: 260 seconds)
17:52:24  * AvianFluquit (Ping timeout: 256 seconds)
17:52:57  * bnoordhuisquit (Ping timeout: 264 seconds)
17:55:19  * hzquit (Ping timeout: 246 seconds)
17:58:26  * defunctzombie_zzchanged nick to defunctzombie
18:01:57  * brsonjoined
18:02:48  <indutny>hoya
18:02:55  <indutny>isaacs: I'm almost done with cpplint
18:04:42  * pachetquit (Quit: leaving)
18:04:46  * amartensjoined
18:05:26  * brsonquit (Client Quit)
18:06:08  * brsonjoined
18:06:55  * TooTallNatequit (Quit: Computer has gone to sleep.)
18:12:18  <MI6>joyent/node: Fedor Indutny master * ccf8f3e : make: add `cpplint` to `test` target (+2 more commits) - http://git.io/SnEc3g
18:12:19  <indutny>tadadadadm
18:22:17  <MI6>nodejs-master: #339 UNSTABLE smartos-x64 (8/617) smartos-ia32 (2/617) http://jenkins.nodejs.org/job/nodejs-master/339/
18:23:46  * AvianFlujoined
18:25:48  <MI6>nodejs-master-windows: #141 FAILURE http://jenkins.nodejs.org/job/nodejs-master-windows/141/
18:27:03  <trevnorris>indutny: any way we can get rid of all the "Done processing src/*" output?
18:31:30  <indutny>hm...
18:33:45  * piscisaureus_joined
18:36:02  <MI6>joyent/node: Fedor Indutny master * 6caf012 : etw_provider: unbreak windows build - http://git.io/asNPBg
18:36:05  <indutny>trevnorris: pull requests are welcome
18:36:13  <trevnorris>...
18:39:30  <trevnorris>tjfontaine: know what's going on with http://jenkins.nodejs.org/job/nodejs-master-windows/141/ ?
18:40:00  * TooTallNatejoined
18:41:25  <trevnorris>and would someone mind fixing the cpp file processing output? don't know python, and don't feel like learning it right now.
18:41:43  <indutny>trevnorris: I think I broke it
18:41:53  <indutny>and probably fixed it
18:46:47  * brsonquit (Quit: Lost terminal)
18:47:12  <MI6>nodejs-master: #340 UNSTABLE smartos-x64 (9/617) smartos-ia32 (3/617) osx-ia32 (1/617) http://jenkins.nodejs.org/job/nodejs-master/340/
18:47:56  * brsonjoined
18:53:08  * dominictarrquit (Quit: dominictarr)
18:54:45  * philips_quit (Read error: Connection reset by peer)
18:59:28  * bnoordhuisjoined
19:00:12  <MI6>nodejs-master-windows: #142 UNSTABLE windows-x64 (16/617) windows-ia32 (18/617) http://jenkins.nodejs.org/job/nodejs-master-windows/142/
19:01:29  <trevnorris>well, just figured out another ridiculous performance trick.
19:03:36  * indexzerojoined
19:04:04  * philips_joined
19:04:22  * bnoordhuisquit (Ping timeout: 256 seconds)
19:05:33  * paulfryz_quit (Remote host closed the connection)
19:06:08  * paulfryzeljoined
19:07:11  * paulfryz_joined
19:10:12  <trevnorris>isaacs: ping
19:10:18  * pachetjoined
19:10:25  * paulfryzelquit (Ping timeout: 245 seconds)
19:15:13  <trevnorris>TooTallNate: found a funny way to store an alloc length in the object:
19:15:25  <TooTallNate>trevnorris: ya?
19:15:29  <trevnorris>TooTallNate: var b = new Number(5); smalloc.alloc(b, b);
19:15:34  <trevnorris>then use b.valueOf();
19:15:41  <TooTallNate>haha
19:15:46  <TooTallNate>nice
19:17:58  <trevnorris>TooTallNate: in 5953 i'm working on passing the array type. something like: smalloc.alloc(8, smalloc.Types.Uint16);
19:18:04  <trevnorris>TooTallNate: like the api?
19:27:21  * bnoordhuisjoined
19:42:47  * mikealquit (Quit: Leaving.)
19:44:24  * AndreasMadsenquit (Remote host closed the connection)
19:51:57  <MI6>joyent/node: Antony Bailey master * 58159e3 : doc: remove travis status image - http://git.io/nf15Aw
19:55:19  * `3E|AFKchanged nick to `3E
19:57:04  <bnoordhuis>$ make test
19:57:05  <bnoordhuis>touch out/Makefile
19:57:05  <bnoordhuis>/usr/bin/python tools/gyp_node -f ninja
19:57:05  <bnoordhuis>ninja -C out/Release/
19:57:05  <bnoordhuis>ninja: Entering directory `out/Release/'
19:57:17  <bnoordhuis>^ seriously, i fucking hate that "feature" of the makefile
19:57:46  * dominictarrjoined
20:02:25  <MI6>nodejs-master: #341 UNSTABLE smartos-x64 (10/617) smartos-ia32 (2/617) osx-ia32 (1/617) http://jenkins.nodejs.org/job/nodejs-master/341/
20:04:22  * pachetquit (Quit: leaving)
20:05:06  <TooTallNate>trevnorris: i mean ya i like the API
20:05:18  <TooTallNate>trevnorris: i could probably figure out a way to use that with "ref"
20:09:02  * paulfryz_quit (Remote host closed the connection)
20:12:55  * Benvie_quit (Ping timeout: 256 seconds)
20:14:42  * pachetjoined
20:15:06  <MI6>nodejs-master-windows: #143 UNSTABLE windows-x64 (18/617) windows-ia32 (18/617) http://jenkins.nodejs.org/job/nodejs-master-windows/143/
20:16:14  * mikealjoined
20:16:28  * Benviejoined
20:17:11  * julianduquejoined
20:20:25  * mikealquit (Ping timeout: 245 seconds)
20:22:34  <indutny>bnoordhuis: heh
20:26:16  * piscisaureus_quit (Ping timeout: 264 seconds)
20:29:19  * julianduquequit (Ping timeout: 246 seconds)
20:29:50  * julian_duquejoined
20:31:05  * hzjoined
20:31:45  * bradleymeckquit (Quit: bradleymeck)
20:32:26  * mikealjoined
20:38:25  * jmar777quit (Remote host closed the connection)
20:41:19  * pachet_joined
20:42:31  * julian_duquequit (Read error: Connection reset by peer)
20:42:42  * julianduquejoined
20:43:20  * pachetquit (Ping timeout: 246 seconds)
20:46:07  * pachet_changed nick to pachet
20:46:12  * pachetquit (Changing host)
20:46:12  * pachetjoined
20:51:59  * julian_duquejoined
20:52:23  * julianduquequit (Ping timeout: 240 seconds)
20:52:35  * TooTallNatequit (Quit: Computer has gone to sleep.)
20:55:51  * rendarquit
21:00:19  * julianduquejoined
21:01:26  * julian_duquequit (Ping timeout: 240 seconds)
21:06:22  * julianduquequit (Ping timeout: 240 seconds)
21:08:12  <trevnorris>bnoordhuis: ping
21:10:28  * hzquit (Disconnected by services)
21:10:32  * hzjoined
21:10:34  * c4milo_quit (Remote host closed the connection)
21:11:01  * c4milojoined
21:12:17  * julianduquejoined
21:15:24  * c4miloquit (Ping timeout: 240 seconds)
21:15:41  * julianduquequit (Client Quit)
21:16:14  <bnoordhuis>trevnorris: pong
21:16:24  * julianduquejoined
21:16:36  <trevnorris>bnoordhuis: in your comment you mention that headers are ISO-8859-1, which I believe is latin1 correct?
21:16:42  * dominictarrquit (Quit: dominictarr)
21:17:37  <trevnorris>bnoordhuis: sorry, ambiguous. here: https://github.com/joyent/node/pull/5961/files#L0R788
21:18:12  <MI6>joyent/node: Ben Noordhuis master * 02cab97 : src: more lint after cpplint tightening (+1 more commits) - http://git.io/sOWPNw
21:18:22  <bnoordhuis>^ i probably broke the windows build :)
21:18:26  <bnoordhuis>trevnorris: yeah, that's correct
21:18:41  <trevnorris>bnoordhuis: since the upgrade to WriteOneByte, binary encoding == latin1 encoding
21:19:00  <trevnorris>bnoordhuis: plus it's significantly faster for large writes.
21:19:27  <trevnorris>faster than ascii encoding that is
21:19:29  <bnoordhuis>is WriteOneByte() latin1 or just 8 bits?
21:20:04  <bnoordhuis>i guess they're pretty much interchangeable
21:20:30  <bnoordhuis>trevnorris: so you're suggesting to switch from ascii to binary?
21:21:04  <trevnorris>bnoordhuis: yes. it'd be faster, at least, and it operates the same if we know if all characters in the string are already in ascii space.
21:21:58  <bnoordhuis>right
21:22:08  <bnoordhuis>btw, that PR is for v0.10 - that doesn't have WriteOneByte()
21:22:12  <trevnorris>bnoordhuis: ah hell, wait.
21:22:22  <trevnorris>i didn't notice your PR is for v0.10
21:23:18  <trevnorris>bnoordhuis: nm. v0.10 doesn't use the new api and binary is a hell a lot slower.
21:23:30  <trevnorris>sorry for wasting your time. :)
21:24:44  <bnoordhuis>heh, no worries
21:26:44  * hzquit (Disconnected by services)
21:26:48  * hzjoined
21:27:12  <trevnorris>bnoordhuis: if this code is still applicable in master then me thinks changing to binary encoding when merged is a good idea.
21:27:59  <bnoordhuis>trevnorris: i think fedor fixed that in master a while ago
21:28:06  <trevnorris>ok cool.
21:28:07  <bnoordhuis>but if he hasn't, then yes, good idea
21:28:38  <bnoordhuis>i just wonder what to do when something actually sends something that's > 127
21:29:14  <bnoordhuis>you just can't win with non-ascii headers. somewhere, something will get terminally confused by it
21:30:28  <trevnorris>note in the docs maybe? I'd like to think it's an edge case that no one would bitch about.
21:32:18  <bnoordhuis>yeah, that's probably a good idea
21:32:21  <MI6>nodejs-master: #342 UNSTABLE smartos-x64 (10/617) smartos-ia32 (2/617) http://jenkins.nodejs.org/job/nodejs-master/342/
21:32:25  <bnoordhuis>"it's not a bug, it's documented!"
21:32:42  <MI6>nodejs-master-windows: #144 FAILURE http://jenkins.nodejs.org/job/nodejs-master-windows/144/
21:32:53  <trevnorris>yeah. and honestly I'd hope we'd never get a ticket about that anyways.
21:33:06  * groundwaterjoined
21:36:27  * julianduquequit (Read error: Connection reset by peer)
21:37:18  <bnoordhuis>isaacs isn't around, is he?
21:37:25  <trevnorris>doesn't seem like it.
21:37:28  * indexzeroquit (Quit: indexzero)
21:37:35  <bnoordhuis>okay, then i'm just going to land the patch
21:38:02  <trevnorris>sounds good. if you can get past the bit hackery, it's pretty straight forward.
21:38:22  * julianduquejoined
21:39:24  <bnoordhuis>i know, right? and if someone gets intimidated by it, more's the pity for him/her
21:39:33  <trevnorris>heh
21:40:20  <bnoordhuis>on the subject of http, does anyone know why we support trailing headers?
21:40:29  <bnoordhuis>i mean, is there anyone who actually uses that?
21:40:59  <trevnorris>if you remove it, i'm sure some random dude will pop up :P
21:41:22  <trevnorris>but not that I know of.
21:41:33  * TooTallNatejoined
21:43:11  <MI6>joyent/node: Ben Noordhuis v0.10 * 3398cce : http: improve chunked res.write(buf) performance - http://git.io/NtYODg
21:45:29  <bnoordhuis>anything i need to look at?
21:45:48  <bnoordhuis>if not, i'm going back to rewriting all the things in haskell
21:46:06  <mmalecki>nice bit hacks right there, Ben
21:46:15  <trevnorris>bnoordhuis: I sent an update to https://github.com/joyent/node/pull/5953
21:46:16  <bnoordhuis>thank you, maciej :)
21:46:28  <trevnorris>bnoordhuis: but it's definitely nothing close to urgent.
21:46:44  <bnoordhuis>looking
21:47:36  <mmalecki>Haskell scares me, maybe because I was using XMonad for some time and when I messed up the syntax, I'd end up with a really lame desktop
21:47:41  <bnoordhuis>trevnorris: what do you think, is Uint8Clamped self-explanatory enough?
21:47:55  <trevnorris>bnoordhuis: yeah. it's the same name as the typed array.
21:48:04  <bnoordhuis>mmalecki: that's what revision control is for right? :)
21:48:08  <trevnorris>bnoordhuis: so either they know it, or google will tell them easily.
21:48:19  <bnoordhuis>trevnorris: yeah, i think i agree
21:48:40  <mmalecki>bnoordhuis: meh, tell that to a 15 yo me
21:50:24  * c4milojoined
21:51:13  <bnoordhuis>ai... there was a fly walking on my laptop's screen
21:51:20  <bnoordhuis>i swatted it without thinking... bad idea
21:51:30  * bnoordhuisgets a napkin
21:51:32  <trevnorris>hehe.
21:54:58  * hzquit
21:55:38  <MI6>nodejs-v0.10: #1387 UNSTABLE smartos-x64 (2/592) osx-ia32 (1/592) http://jenkins.nodejs.org/job/nodejs-v0.10/1387/
21:56:38  * pachetquit (Ping timeout: 240 seconds)
21:57:18  <bnoordhuis>trevnorris: oh btw, make sure it passes `make cpplint`
21:57:41  <trevnorris>bnoordhuis: ah yeah. i'll rebase on master. thanks.
21:58:23  <bnoordhuis>mmalecki: re haskell, it has some downsides but it's got the nicest parser libraries that i know of
21:58:29  <trevnorris>bnoordhuis: wow, I actually fixed that correctly on my first round. :)
21:59:33  <trevnorris>bnoordhuis: re macros: I've been trying to ping isaacs about that one. should be IS_PRIMITIVE(arg) = (!((arg) instanceof Object))
21:59:45  * pachetjoined
21:59:48  <trevnorris>but didn't want to add it since the other is in the works.
22:00:01  <bnoordhuis>trevnorris: don't know if there's much point, we're going to roll it back, right?
22:01:16  * philips_quit (Changing host)
22:01:16  * philips_joined
22:03:04  <trevnorris>bnoordhuis: for now, but i'm wouldn't be surprised if that changed.
22:05:42  <MI6>nodejs-v0.10-windows: #121 UNSTABLE windows-x64 (9/592) windows-ia32 (10/592) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/121/
22:05:57  <bnoordhuis>what do you know that i don't? :)
22:07:07  <trevnorris>nothing. i just think it's a topic worth revisiting down the road.
22:07:27  <bnoordhuis>yeah, me too
22:09:34  * indexzerojoined
22:14:29  <trevnorris>bnoordhuis: so I added using_domains to node_internals.h, which really helps performance. think we could add bool node::DomainsInUse() or some such to node.h?
22:14:42  <trevnorris>checking if a domain object is setup on the global context is painfully slow
22:15:06  <bnoordhuis>trevnorris: why would you want to export that?
22:15:32  <trevnorris>bnoordhuis: users should always use MakeCallback when reentering JS from C++ after an async operation.
22:16:04  <bnoordhuis>they should use MakeCallback(), period :)
22:16:05  <trevnorris>bnoordhuis: and for proper use of domains they need to check if they're in use at call time.
22:16:20  <trevnorris>that doesn't solve the issue with domains.
22:16:30  <bnoordhuis>why? MakeCallback() takes care of that, right?
22:16:31  * mikealquit (Quit: Leaving.)
22:16:44  <trevnorris>it'll enter the domain, but the domain object needs to be attached to the request
22:16:53  * mikealjoined
22:16:53  <trevnorris>or the domain can't know it's tracking the same async call
22:17:23  <bnoordhuis>i should hope that add-on authors that use domains have a js shim that sets up the object
22:17:25  <trevnorris>users can have multiple domains in a script, so we need to attached the request specific domain when the function is run.
22:18:27  <bnoordhuis>adding a new api function isn't the worst thing but it doesn't really seem necessary in this case
22:19:36  <trevnorris>it's do-able w/o the api, but much slower. either create the object in js then pass it around in c++ or they check if the domain is set on the global context right before the function is run.
22:20:00  <trevnorris>imho the first is over complicated and the second is slow.
22:22:02  <trevnorris>there could totally be another option that I'm not seeing, but the second example is what's happening in req_wrap.h in v0.10
22:22:29  <trevnorris>not friendly. :(
22:22:35  <bnoordhuis>but, but... the 'create in js and pass to c++' thing is what _we_ do, right?
22:22:43  <bnoordhuis>all the setup and teardown is done in js land
22:24:01  <trevnorris>heh, only when it's faster :)
22:25:18  <bnoordhuis>another thing, that DomainsInUse() function only returns true when something is already using domains, right?
22:25:41  <trevnorris>it will only return true if the domain module has been loaded.
22:25:53  <bnoordhuis>so if you're an add-on author and your add-on is the first thing to load, it won't help you determine if domains should actually be used
22:26:12  <bnoordhuis>chicken/egg problem
22:26:35  * AvianFlu_joined
22:26:57  * AvianFlu_quit (Remote host closed the connection)
22:27:04  <trevnorris>DomainInUse() would be called every time right before an async method is run from c++ that has a JS callback.
22:27:15  <trevnorris>take a look at req_wrap.h on master.
22:27:36  <trevnorris>it checks using_domains on every instantiation to see if the domain module has been loaded yet.
22:27:55  <trevnorris>if it hasn't then domains can't be in use, so we don't bother checking the global context for a domain object.
22:28:18  <bnoordhuis>yeah, i get that
22:28:31  <bnoordhuis>have people been asking for this?
22:28:52  <trevnorris>they don't know they should be honestly. this came up from reviewing https://github.com/joyent/node/pull/5809
22:29:08  <trevnorris>he's calling the js callback right from the c++ callback.
22:29:10  * julianduquequit (Ping timeout: 245 seconds)
22:29:24  <trevnorris>i'm right in the middle of writing a massive response on how to properly use MakeCallback with domains.
22:29:33  <bnoordhuis>with fun->Call(), you mean?
22:29:36  <trevnorris>yeah.
22:29:43  <bnoordhuis>right
22:29:49  <trevnorris>that'll fail if domains are in use, and it won't resolve the nextTickQueue
22:30:00  * AvianFluquit (Ping timeout: 245 seconds)
22:30:39  <bnoordhuis>well, i don't know about fail - it just skips all the domains logic
22:30:51  <bnoordhuis>some people probably feel that's a good thing
22:30:55  <trevnorris>haha.
22:31:00  * trevnorrisraises hand
22:31:19  <trevnorris>but not processing any nextTick callbacks could be fatal
22:31:26  <trevnorris>not likely, but there's still a small could
22:31:33  <bnoordhuis>yeah, that's true
22:31:54  <bnoordhuis>re domains, i've had several people tell me in the last few weeks how much they hate it from a performance perspective
22:32:17  <trevnorris>yeah. and i've already spent a ridiculous amount of time performance tuning those things.
22:32:21  <bnoordhuis>if you have an app where a single module uses domains somewhere, your entire app pays the performance penalty
22:32:25  <bnoordhuis>yeah, good work, trevor
22:32:38  <trevnorris>I understand the issue, but I'm not sure how else to get around it.
22:32:56  <trevnorris>wait. I might have another ridiculous performance idea.
22:33:01  <bnoordhuis>oh dear
22:33:15  <trevnorris>oh hells yeah. i'm going to work on that soon.
22:33:18  * julianduquejoined
22:33:23  <trevnorris>hopefully that'll fix the issue.
22:33:29  <bnoordhuis>i'm afraid to ask
22:34:00  <trevnorris>haha. it's not unlike the infoBox hack.
22:34:22  <bnoordhuis>using a typed array somewhere?
22:35:29  <trevnorris>yeah. the idea is to toggle state when domains are actively being used. instead of just, well domains are loaded so they're always in use.
22:36:14  <trevnorris>though i'm not sure how this'll work with your multi context changes. i'm interested to see if those have applications in core.
22:36:50  <bnoordhuis>it shouldn't have much impact
22:37:06  <bnoordhuis>the most visible change in the code base is that some static variables move to a per-context structure
22:37:06  <trevnorris>ok cool. then I won't worry about it.
22:37:41  <trevnorris>ok, would this be like a more "true" vm?
22:38:06  <bnoordhuis>it's one step closer to full isolate support, if that is what you mean
22:38:16  <trevnorris>coolio.
22:38:57  <bnoordhuis>i guess i should rebase my multi-context stuff against master
22:39:01  <bnoordhuis>not looking forward to that...
22:39:09  <trevnorris>been a while since you have?
22:39:21  <bnoordhuis>yep. two, maybe three months
22:39:29  <trevnorris>ooooh. yeah. good luck.
22:39:43  <bnoordhuis>thanks :)
22:40:03  <bnoordhuis>okay, signing off. have a fun evening, trevor
22:40:47  <trevnorris>you to ben. night.
22:40:52  <trevnorris>*too
22:44:10  * bnoordhuisquit (Ping timeout: 245 seconds)
22:49:10  * stagasjoined
22:50:07  * c4miloquit (Remote host closed the connection)
22:50:34  * c4milojoined
22:54:01  * pachetquit (Quit: leaving)
22:55:02  * c4miloquit (Ping timeout: 240 seconds)
22:56:31  * indexzeroquit (Quit: indexzero)
23:09:27  * dominictarrjoined
23:39:46  <tjfontaine>trevnorris: what do you mean what's up? indutny broke the build :)
23:39:55  <trevnorris>tjfontaine: haha
23:41:36  * mikealquit (Quit: Leaving.)
23:42:17  <tjfontaine>indutny: https://gist.github.com/tjfontaine/6127247 please fix the windows build.
23:45:36  <trevnorris>TooTallNate: well, if you want to play around w/ it, it's been signed off by ben (so probably won't crash your machine ;) https://github.com/joyent/node/pull/5953
23:46:24  * felixgequit (Quit: felixge)
23:49:55  * dominictarrquit (Quit: dominictarr)
23:50:38  * bnoordhuisjoined
23:52:25  * toothrchanged nick to toothrot
23:54:45  * bnoordhuisquit (Ping timeout: 248 seconds)
23:59:25  * brianc_quit (Remote host closed the connection)