00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:09  * ircretaryjoined
00:05:46  <isaacs>i think i'm gonna add a flag for that, i guess.
00:05:52  <isaacs>but we shouldn't fix on 0.10
00:05:56  <isaacs>just leave it as-is
00:08:14  <MI6>nodejs-master-windows: #4 UNSTABLE windows-x64 (9/587) windows-ia32 (9/587) http://jenkins.nodejs.org/job/nodejs-master-windows/4/
00:08:17  <tjfontaine>that's difficult ya
00:08:27  <tjfontaine>oh
00:08:37  <tjfontaine>if windows finished and I didn't see nodejs-master that means ...
00:08:45  <tjfontaine>LOUDBOT: HOLY SHIT GREEN BUILD
00:08:45  <LOUDBOT>tjfontaine: STOP BEING A BAD CATALYST :<
00:09:05  <tjfontaine>oh sadly not
00:09:23  <tjfontaine>just a dropped packet somewhere
00:09:35  * dominictarrquit (Quit: dominictarr)
00:12:18  * st_lukejoined
00:20:28  <toothrot>dang... still no subdirectory watching.
00:21:10  <isaacs>tjfontaine: that's what you get for being a bad catalyst
00:21:19  <tjfontaine>I know :/
00:21:26  <tjfontaine>loudbot is always on point
00:22:16  <toothrot>where the heck is the search box for issues
00:22:24  * qardquit (Quit: Leaving.)
00:22:26  <tjfontaine>they fucking [re]moved it
00:23:14  <toothrot>sheesh.
00:23:21  <toothrot>YEAH I SAID IT
00:23:21  <LOUDBOT>CAN'T BE INVISIBLE IF YOU'RE UPSIDE DOWN, SILLY
00:24:22  <tjfontaine>toothrot: in short they integrated it into the top search bar
00:25:55  <toothrot>i don't see that... need to be logged in maybe?
00:26:18  <tjfontaine>ah maybe
00:26:25  <tjfontaine>I don't see it in incognito
00:26:44  <toothrot>ah, well i can log in then... thanks for the tip
00:27:53  * c4miloquit (Remote host closed the connection)
00:31:53  * amartensquit (Quit: Leaving.)
00:32:25  * c4milojoined
00:33:52  <tjfontaine>that one test fix is hilarious: (18 failures / -275)
00:42:16  <isaacs>hmm...
00:42:37  <isaacs>i think i'm leaning more towards "don't care" about the write() cb getting called before the 'finish' event
00:44:50  * mikealquit (Quit: Leaving.)
00:45:33  <tjfontaine>I don't have a good scenario on why I would care, but it feels like we should care
00:48:37  <isaacs>i don't think anyone cares today
00:48:41  <isaacs>since no one's complaining about it
00:49:18  <isaacs>except... we DO guarantee that the write() cb is called asynchronously, but we DONT guarantee that end() emits 'finish' asynchronously
00:50:00  <isaacs>and it makes Transform really tricky to know when to finish up, if we make finish async
00:50:12  <tjfontaine>nod
00:50:23  <isaacs>becaues it's hanging on that hook
00:55:09  * CoverSli1equit (Read error: Operation timed out)
00:56:31  <isaacs>yeah, i'm preemptively closing this bug WONTFIX
00:56:43  <isaacs>and i'll come up with a better way to test writev :)
00:57:41  * CoverSlidejoined
01:15:15  * inolenquit (Quit: Leaving.)
01:17:13  * dapquit (Quit: Leaving.)
01:36:23  * abraxasjoined
01:58:26  * kenperkinsquit (Read error: Connection reset by peer)
01:58:37  * Correnjoined
02:09:00  * brsonquit (Ping timeout: 258 seconds)
02:13:25  <MI6>libuv-v0.10: #52 UNSTABLE smartos (3/186) windows (5/187) linux (1/186) osx (1/187) http://jenkins.nodejs.org/job/libuv-v0.10/52/
02:19:41  * qardjoined
02:21:05  * Correnchanged nick to kenperkins
02:27:20  * inolenjoined
02:28:39  * c4miloquit (Read error: Connection reset by peer)
02:31:37  * c4milojoined
02:31:51  * qardquit (Quit: Leaving.)
02:50:08  * AvianFlujoined
03:08:22  * defunctzombie_zzchanged nick to defunctzombie
03:12:47  * c4miloquit (Remote host closed the connection)
03:14:46  * AvianFluquit (Read error: Connection reset by peer)
03:15:39  * AvianFlujoined
03:17:08  * brsonjoined
03:41:37  <MI6>libuv-node-integration: #43 ABORTED osx-x64 (1/582) osx-ia32 (1/582) windows-ia32 (9/582) smartos-x64 (1/582) windows-x64 (12/582) http://jenkins.nodejs.org/job/libuv-node-integration/43/
03:45:41  * defunctzombiechanged nick to defunctzombie_zz
04:23:02  * Correnjoined
04:24:17  * kenperkinsquit (Ping timeout: 268 seconds)
04:25:36  * Correnchanged nick to kenperkins
04:27:46  * normanmjoined
04:55:34  * defunctzombie_zzchanged nick to defunctzombie
05:05:05  * defunctzombiechanged nick to defunctzombie_zz
05:51:46  * st_lukequit (Remote host closed the connection)
06:02:53  * rendarjoined
06:06:30  * timoxleyquit (Quit: Computer has gone to sleep.)
06:09:46  * bnoordhuisjoined
06:10:44  * st_lukejoined
06:15:32  * stolsmajoined
06:18:13  * Correnjoined
06:22:23  * kenperkinsquit (Ping timeout: 255 seconds)
06:24:14  * wolfeida_joined
06:25:28  * wolfeidauquit (Ping timeout: 245 seconds)
06:30:27  * mikealjoined
06:53:51  * csaohjoined
06:55:24  * brsonquit (Quit: leaving)
06:57:04  * csaohquit (Client Quit)
06:58:27  * bnoordhuisquit (Ping timeout: 268 seconds)
07:01:52  * AvianFluquit (Read error: Connection reset by peer)
07:02:27  * AvianFlujoined
07:21:20  * loladiroquit (Quit: loladiro)
07:25:31  * `3rdEdenjoined
07:30:03  * st_lukequit (Remote host closed the connection)
07:34:06  * loladirojoined
07:36:46  * csaohjoined
07:36:48  * csaohquit (Remote host closed the connection)
07:38:03  * stagasjoined
07:40:46  * csaohjoined
07:41:09  * timoxleyjoined
07:44:42  * loladiroquit (Quit: loladiro)
07:52:06  * AvianFluquit (Remote host closed the connection)
07:54:06  * stolsmaquit (Ping timeout: 272 seconds)
08:11:05  * wolfeida_changed nick to wolfeidau
08:27:10  * Correnquit (Read error: Connection reset by peer)
08:28:01  * kenperkinsjoined
08:34:28  * stagasquit (Quit: ChatZilla 0.9.90-rdmsoft [XULRunner 1.9.0.17/2009122204])
08:52:22  * bnoordhuisjoined
09:02:28  * bnoordhuisquit (Quit: leaving)
09:07:51  * mraleph1quit (Read error: Connection reset by peer)
09:07:53  * mralephjoined
09:20:03  * icarotjoined
09:32:48  * c4milojoined
09:34:12  * hzjoined
09:39:35  * icarotquit (Quit: leave())
09:40:06  * icarotjoined
09:42:42  * raljoined
09:48:35  * benoitcquit (Ping timeout: 246 seconds)
10:12:29  * c4miloquit (Remote host closed the connection)
10:19:29  * kazuponjoined
10:39:32  * icarotquit (Remote host closed the connection)
11:01:10  * benoitcjoined
11:02:10  * kazuponquit (Remote host closed the connection)
11:11:46  * abraxasquit (Remote host closed the connection)
11:24:33  * stolsmajoined
12:18:56  * timoxleyquit (Quit: Computer has gone to sleep.)
13:02:44  * kazuponjoined
13:11:32  * kazuponquit (Ping timeout: 268 seconds)
13:12:02  * abraxasjoined
13:13:01  * timoxleyjoined
13:16:23  * abraxasquit (Ping timeout: 245 seconds)
13:28:45  * defunctzombie_zzchanged nick to defunctzombie
13:40:40  * indexzerojoined
13:42:51  * pachetjoined
13:43:29  * AvianFlujoined
13:56:19  * indexzeroquit (Quit: indexzero)
13:56:54  * kevinswiberjoined
13:57:49  * loladirojoined
14:07:33  * c4milojoined
14:07:44  * kazuponjoined
14:15:04  * stagasjoined
14:15:59  * kazuponquit (Ping timeout: 260 seconds)
14:30:38  * c4miloquit (Remote host closed the connection)
14:37:06  * mikealquit (Quit: Leaving.)
14:49:00  * stagasquit (Read error: Connection reset by peer)
14:51:16  * c4milojoined
15:07:46  * indexzerojoined
15:12:53  * kazuponjoined
15:14:18  * timoxleyquit (Quit: Computer has gone to sleep.)
15:16:33  * bnoordhuisjoined
15:17:42  * kazuponquit (Ping timeout: 252 seconds)
15:24:37  * `3rdEdenquit (Remote host closed the connection)
15:25:40  * indexzeroquit (Quit: indexzero)
15:26:18  * dominictarrjoined
15:40:10  * dapjoined
15:42:35  * piscisaureus_joined
15:43:32  * dominictarrquit (Quit: dominictarr)
15:47:44  * stagasjoined
15:48:25  * loladiroquit (Quit: loladiro)
15:48:37  <bnoordhuis>call in 10?
15:53:28  <tjfontaine>afaik
15:53:35  <tjfontaine>at least isaacs said yes yesterday
15:54:00  <bnoordhuis>i'm ready, willing and able
15:54:21  <tjfontaine>I figured you would be traveling
15:54:39  <bnoordhuis>no, we arrived yesterday
15:54:51  <tjfontaine>ah welcome
15:54:55  <bnoordhuis>thank you :)
15:55:32  <tjfontaine>:)
15:56:21  <trevnorris>bnoordhuis: so the meetup on thurs is full. can I still attend?
15:56:33  <tjfontaine>I'm crashing it :P
15:56:37  <trevnorris>heh
15:57:57  <bnoordhuis>trevnorris: should be no problem. we reserved a couple of seats for people like tjfontaine
15:58:49  <tjfontaine>isaacs in the office now, so we'll be starting the call shortly
15:59:00  <bnoordhuis>okay, cool
15:59:24  * dominictarrjoined
16:00:48  <piscisaureus_>I'll join too
16:03:19  * defunctzombiechanged nick to defunctzombie_zz
16:04:21  <bnoordhuis>indutny: call
16:06:17  * mikealjoined
16:13:20  * TooTallNatejoined
16:13:31  * kazuponjoined
16:13:34  * brsonjoined
16:15:47  * Correnjoined
16:18:31  * kenperkinsquit (Ping timeout: 276 seconds)
16:18:31  * kazuponquit (Ping timeout: 264 seconds)
16:18:48  * Correnquit (Read error: Connection reset by peer)
16:18:58  * kenperkinsjoined
16:21:00  * mikealquit (Quit: Leaving.)
16:25:46  * kenperkinschanged nick to ken
16:25:56  * kenchanged nick to kenperkins
16:28:56  * Correnjoined
16:29:48  * Benvie_quit (Read error: Connection reset by peer)
16:29:49  * loladirojoined
16:30:04  * Benvie_joined
16:30:57  * qardjoined
16:31:20  * kenperkinsquit (Ping timeout: 268 seconds)
16:33:05  <indutny>oh
16:33:06  <indutny>sorry
16:33:07  <indutny>was afk
16:35:14  <tjfontaine>call is still going at the moment
16:36:13  <indutny>oh really?
16:37:20  * kevinswiberquit (Remote host closed the connection)
16:38:04  * kevinswiberjoined
16:42:08  * TooTallNatequit (Quit: Computer has gone to sleep.)
16:44:19  * bnoordhuisquit (Ping timeout: 264 seconds)
16:45:10  * amartensjoined
16:47:55  * inolenquit (Quit: Leaving.)
16:48:42  * piscisaureus_quit (Ping timeout: 264 seconds)
16:52:59  <trevnorris>isaacs: here's another change on the roadmap: "Changing all callbacks that return Handle<Value> to return void and instead pass the return value in as a parameter."
16:53:43  <tjfontaine>that's going to be painful of course :)
16:54:09  <trevnorris>yup. and it just got a "lgtm" today: https://codereview.chromium.org/12494012/'
16:56:50  * stagasquit (Read error: Connection reset by peer)
16:56:57  <trevnorris>oh son of a bitch. so Persistents no longer inherit from Handle.
16:57:19  <trevnorris>before you could do a "FunctionDeclaration(Handle<Value>)" and it would accept a persistent or local
16:58:22  * dominictarrquit (Ping timeout: 245 seconds)
16:58:41  * dominictarrjoined
16:59:28  * benoitcquit (Excess Flood)
17:00:05  * TooTallNatejoined
17:01:10  <trevnorris>TooTallNate: in advance, sorry.
17:01:19  <TooTallNate>lol
17:01:20  <TooTallNate>ok
17:01:37  <trevnorris>v8 3.19 is going to require possibly major refactoring.
17:01:52  <TooTallNate>in what?
17:02:12  <trevnorris>hm. Local<> is going to be deprecated.
17:02:24  <isaacs>trevnorris: oh, holy hell.
17:02:38  <TooTallNate>trevnorris: what replaces it?
17:02:41  <trevnorris>yeah. i'm looking at the patches now.
17:02:45  <trevnorris>TooTallNate: Handle<>
17:02:54  <trevnorris>https://codereview.chromium.org/12729023
17:02:56  <isaacs>trevnorris: so, like, instead of Local<String> s = String::New("asdf"); it'll be Handle<String> s; String::New("asdf", s);?
17:03:01  <trevnorris>https://codereview.chromium.org/12494012/
17:03:04  <trevnorris>TooTallNate: yeah
17:03:26  <TooTallNate>why don't they just leave Local hanging around for backwards compat? :p
17:03:43  <trevnorris>they are, under the "UNSAFE_HANDLES" flag at compile time.
17:03:44  <qard>Horribly breaking changes, yay! :O
17:04:05  * csaohquit (Quit: csaoh)
17:04:42  <TooTallNate>well we'll have to at least add a compile-time warning for native addons to have time to upgrade
17:04:55  <trevnorris>TooTallNate: also, things that returned Handle<Value> now return void, and the return value is a fn parameter.
17:05:10  <TooTallNate>interesting
17:05:14  <TooTallNate>that's a bigger change
17:05:34  <TooTallNate>trevnorris: is this in the name of speed/optimizations in V8?
17:05:40  * benoitcjoined
17:05:59  <trevnorris>TooTallNate: yeah. Persistents were horribly slow. this should provide substantial improvements.
17:06:10  <trevnorris>but at the expense of backwards compatibility. =P
17:06:11  <TooTallNate>trevnorris: so Persistent is gone now as well?
17:06:30  <trevnorris>TooTallNate: oh, no. it's still around.
17:06:38  <trevnorris>just doesn't inherit from Handle<>
17:06:58  <trevnorris>TooTallNate: so before you could "Fn(Handle<Value> arg)" and it would accept Persistent or Local
17:07:02  <trevnorris>but now that's not the case.
17:07:07  <TooTallNate>right, right
17:07:17  <TooTallNate>well… that's annoying :p
17:07:22  <trevnorris>yeah.
17:08:24  <trevnorris>as soon as the bigger of the two changesets is committed, going to see what happens when upgrading node to bleeding_edge
17:08:50  <trevnorris>isaacs: oh, yeah. (responded to TooTallNate)
17:10:38  * bajtosjoined
17:11:23  * mikealjoined
17:12:39  * abraxasjoined
17:14:10  * kazuponjoined
17:17:40  * abraxasquit (Ping timeout: 276 seconds)
17:18:36  * kazuponquit (Ping timeout: 256 seconds)
17:29:07  * inolenjoined
17:31:18  * icarotjoined
17:35:24  * Correnchanged nick to kenperkins
17:59:13  * bnoordhuisjoined
18:00:40  <isaacs>trevnorris: this V8 update is gonna suuuuuck
18:01:26  * perezdquit (Quit: perezd)
18:01:40  <isaacs>trevnorris: it seems like a mostly unnecessary change. i mean, C++ has function overloading, why not just have it return a handle if you don't pass one in?
18:01:44  * piscisaureus_joined
18:01:47  * isaacsis doing exactly what is annoying when node users do it...
18:01:52  * bajtosquit (Quit: bajtos)
18:01:54  <tjfontaine>heh
18:02:17  <isaacs>rar! you are changing a thing! i don't understand the reason so therefor you have no reason to make this change! i am angry!!
18:02:23  <trevnorris>lol
18:02:33  <trevnorris>but yeah. this is going to suck.
18:02:41  * bajtosjoined
18:02:46  <trevnorris>and it's going to kill the community.
18:02:53  <trevnorris>(kill as in piss off)
18:03:09  <tjfontaine>ok I was going to say that was a bit harsh :)
18:03:55  <trevnorris>bnoordhuis: when you have a minute: http://git.io/zZJ3pw
18:04:07  * mikealquit (Quit: Leaving.)
18:04:23  <trevnorris>bnoordhuis: if I don't explicitely static_cast to void* it's like the entire buffer is copied into the v8 heap.
18:05:21  <trevnorris>isaacs: funny how we're trying our hardest to stabilize api, but v8's like. eh, what the hell. faster!!!
18:05:59  <bnoordhuis>trevnorris: i'm afraid it's all meetings for me today :(
18:06:17  <trevnorris>bnoordhuis: heh. well I feel worse for you then.
18:06:39  <tjfontaine>sigh, debugger tests are hanging on windows again, blah
18:07:46  <amartens>wow, that's quite the changes to V8
18:10:52  * geNAZt-laptopjoined
18:11:16  <geNAZt-laptop>hi
18:11:35  <geNAZt-laptop>i am using nodejs and i want to spin up a second event loop thread. is this possible ?
18:12:10  <tjfontaine>yes it's possible, why do you need it?
18:12:38  <tjfontaine>or rather, how do you plan on working with it
18:12:49  <geNAZt-laptop>well i will load of heavy calculations
18:13:05  <tjfontaine>why not just use uv_queue_work then?
18:13:10  <geNAZt-laptop>i will use it like libuv handles i/o but for calculations
18:13:22  <geNAZt-laptop>does it work in an extra thread ?
18:13:26  <tjfontaine>yes
18:13:30  <geNAZt-laptop>without blocking the default loop ?
18:13:32  <tjfontaine>yes
18:13:37  <geNAZt-laptop>nice thanks
18:13:41  <tjfontaine>you're welcome
18:13:45  * benoitcquit (Excess Flood)
18:14:40  * defunctzombie_zzchanged nick to defunctzombie
18:14:44  * kazuponjoined
18:19:12  * kazuponquit (Ping timeout: 245 seconds)
18:20:16  * geNAZt-laptoppart ("Verlassend")
18:21:10  * benoitcjoined
18:32:41  * defunctzombiechanged nick to defunctzombie_zz
18:39:08  * Siphonblastjoined
18:41:27  * icarotquit (Ping timeout: 268 seconds)
18:47:49  * csaohjoined
18:50:00  * csaohquit (Client Quit)
18:53:03  * benoitcquit (Excess Flood)
18:54:39  * benoitcjoined
18:54:45  * defunctzombie_zzchanged nick to defunctzombie
19:06:41  * bajtosquit (Quit: bajtos)
19:07:18  * bajtosjoined
19:12:32  * bnoordhuisquit (Ping timeout: 245 seconds)
19:12:47  * piscisaureus_quit (Ping timeout: 255 seconds)
19:15:21  * kazuponjoined
19:20:15  * kazuponquit (Ping timeout: 252 seconds)
19:35:24  * kevinswiberchanged nick to kevinswiber-away
19:39:03  <trevnorris>indutny: I see WriteBuffer arguments weren't updated even though the first argument is no longer needed.
19:39:09  <trevnorris>indutny: is that for api compatibility?
19:43:55  * dominictarrquit (Quit: dominictarr)
19:44:57  * normanmquit (Quit: Computer has gone to sleep.)
19:46:42  * dominictarrjoined
19:48:45  * `3rdEdenjoined
19:52:23  * bnoordhuisjoined
19:58:02  * defunctzombiechanged nick to defunctzombie_zz
20:01:47  <tjfontaine>isaacs: http://jenkins.nodejs.org/job/node-pullrequest/DESTCPU=ia32,label=linux/lastCompletedBuild/tapTestReport/ duration is sortable, so you can see which tests are taking the longest
20:04:13  * kevinswiber-awayquit (Remote host closed the connection)
20:07:26  <tjfontaine>isaacs: for when you have time https://github.com/joyent/node/pull/5419
20:08:46  <trevnorris>tjfontaine: - test-domain-implicit-fs.js 57 sec. seriously?
20:09:41  <tjfontaine>trevnorris: it's possible, anything can happen on a virtualized slave
20:09:50  <trevnorris>ah, I forget about that.
20:12:43  <MI6>joyent/node: Timothy J Fontaine v0.10 * 716176f : test: report test duration in TAP progress - http://git.io/XF0Jrg
20:14:08  <MI6>joyent/node: Timothy J Fontaine master * ee4f0ba : test: report test duration in TAP progress - http://git.io/_nn2_A
20:14:33  <trevnorris>tjfontaine: you still look like your linkedin pic?
20:14:54  <tjfontaine>hmm
20:15:08  <tjfontaine>shorter hair and beard, but sure
20:15:22  * stagasjoined
20:15:56  <indutny>trevnorris: I'm not sure :)
20:15:56  * kazuponjoined
20:16:55  <trevnorris>indutny: ok. i'm mapping out all the *_wrap's.
20:17:03  * kevinswiber-awayjoined
20:17:05  <trevnorris>and just noticed that.
20:17:59  <tjfontaine>github search bar is really awesome at [re]autocomplete https://dsz91cxz97a03.cloudfront.net/gOv4cy-DhK-900x900.jpeg
20:18:37  <amartens>it really wants to search for joyent/node
20:19:11  <tjfontaine>well, start typing, press down, it fills that out, press down, rinse repeat
20:20:55  * kazuponquit (Ping timeout: 264 seconds)
20:24:30  <MI6>nodejs-v0.10: #175 UNSTABLE osx-x64 (1/582) linux-x64 (2/582) linux-ia32 (5/582) http://jenkins.nodejs.org/job/nodejs-v0.10/175/
20:24:53  <tjfontaine>5 failing tests linux? really...
20:24:58  <tjfontaine>oh debugger
20:25:06  <MI6>nodejs-master: #197 UNSTABLE linux-ia32 (1/587) osx-ia32 (1/587) http://jenkins.nodejs.org/job/nodejs-master/197/
20:26:32  * kevinswiber-awayquit (Remote host closed the connection)
20:29:12  <trevnorris>why isn't this in an "#ifdef"? : http://git.io/7ex-dQ
20:29:48  <tjfontaine>because you may pass in something different at runtime?
20:30:46  <trevnorris>tjfontaine: that's win32 specific
20:31:00  <trevnorris>defined in node_win32_perfctr_provider.cc
20:31:16  <tjfontaine>the macro for NODE_COUNT_NET_BYTES_RECV you mean?
20:31:20  <trevnorris>yeah
20:31:49  <trevnorris>the if/else block should be wrapped in a ifdef for windows builds.
20:32:08  <tjfontaine>well "ok" that all perhaps should be hidden, but the macro is pointless on other systems so the compiler probably wipes it all away anyway
20:32:42  <txdv>is it possible to move FDs from one loop to another?
20:32:46  <trevnorris>i like to explicitely know when I'm reading the code that it's win32 specific. instead of spending 20 minutes tracking down what it does.
20:33:23  * bnoordhuisquit (Ping timeout: 245 seconds)
20:33:31  <tjfontaine>trevnorris: I wouldn't call it win32 specific, but perf specific, there's an analog for the dtrace/systemtap stuff as well
20:33:51  <trevnorris>tjfontaine: is that actually injected into the code like those macros?
20:33:56  <tjfontaine>trevnorris: fwiw the first place you should have looked was in node_counters.h
20:34:39  <trevnorris>tjfontaine: and how am I supposed to know that if i've never read the code base before? node_counters was introduced w/ these changes.
20:34:57  <trevnorris>tjfontaine: and is there code injected into core for dtrace?
20:34:57  <tjfontaine>trevnorris: I'm just going by what I saw when I grep'd
20:35:20  <tjfontaine>trevnorris: "injected"? not so much but there are certainly macros that behave the same
20:36:03  <tjfontaine>and as things further solidify more probe points will be added
20:36:25  <tjfontaine>I have some that I've not really done too much with atm, specifically to do with the buffer alloc/free
20:36:30  <trevnorris>tjfontaine: can you give me an example macro for dtrace? all I see are the *.d files.
20:36:55  <tjfontaine>node_dtrace.cc NODE_GC_START|DONE
20:38:50  <trevnorris>tjfontaine: I can't find any of those macros being used in node core. so you're just saying they exist but aren't being used?
20:38:54  <tjfontaine>I also have in my consolidation branch a place to group these macros together, but you certainly did point me towards new ones for the ETW things
20:39:20  <tjfontaine>trevnorris: actually they're used in lib/ :)
20:39:36  <tjfontaine>grep -nr DTRACE_ lib/*
20:39:59  * kenperkinsquit (Read error: Connection reset by peer)
20:40:14  <tjfontaine>but they ultimately fall down into a C macro that's created from a `dtrace` invocation at build time (or an empty macro if dtrace is disabled)
20:40:45  <trevnorris>wait. we're making calls to cc, even if those aren't being used?
20:41:06  * kenperkinsjoined
20:41:17  <tjfontaine>potentially
20:41:25  <trevnorris>...
20:41:30  * bnoordhuisjoined
20:43:05  <tjfontaine>trevnorris: if !dtrace,!systemtap,!etw is available at build time, then they're noop js calls
20:43:41  <trevnorris>tjfontaine: how? they're not a built-in. so there's no such thing as a noop from js to cc
20:43:54  <tjfontaine>trevnorris: see src/macros.py
20:44:39  <trevnorris>tjfontaine: but that doen't remove the COUNTER_HTTP_SERVER_RESPONSE also in lib/
20:44:52  <trevnorris>s/COUNTER_*
20:45:05  <tjfontaine>see also src/perfctr_macros.py
20:51:04  * stolsmaquit (Ping timeout: 256 seconds)
20:55:28  * piscisaureus_joined
20:57:14  <trevnorris>note: I just noticed that "test/fixtures/child-process-persistent.js" has been running in the background since last week.
20:57:29  <tjfontaine>heh
21:00:14  <trevnorris>tjfontaine: so it's normal to have perf metrics code littered through out core code?
21:00:50  <tjfontaine>it is normal to see the probe points show up beside important code ya
21:01:42  * Siphonblastquit (Ping timeout: 245 seconds)
21:05:18  * piscisaureus_quit (Ping timeout: 252 seconds)
21:06:49  <trevnorris>indutny: in stream_wrap it looks like a new WriteWrap is created to attach the request instance to the associated js object instance. is that about correct?
21:07:17  <indutny>em… not exactly
21:07:22  <indutny>it also holds all data
21:07:39  <indutny>but I don't remember details, so probably you're right
21:08:00  <indutny>ah
21:08:03  <indutny>no, you ain't right
21:08:07  <indutny>its holding write request
21:08:10  <MI6>nodejs-v0.10-windows: #6 UNSTABLE windows-ia32 (7/582) windows-x64 (8/582) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/6/
21:08:10  <indutny>and all data
21:09:47  * piscisaureus_joined
21:12:41  <trevnorris>indutny: ah hell, there are so many wrappers there.
21:13:54  <indutny>yeah
21:17:01  * kazuponjoined
21:18:47  <trevnorris>indutny: btw, impressive stuff w/ writev.
21:21:39  * kazuponquit (Ping timeout: 252 seconds)
21:22:27  <indutny>thanks
21:23:10  <trevnorris>indutny: had an idea for that, think my thinbuffer might be useful and want your opinion.
21:23:37  <indutny>well, I've no opinion until you'll explain your thoughts :)
21:23:47  <trevnorris>heh, in process. ;-)
21:24:36  <trevnorris>i'm thinking that people mix/match strings and buffers not very often.
21:25:11  <trevnorris>so create a, um. 10MB allocation and write strings to it as they come in. if that fills then create a new one.
21:25:23  <trevnorris>then when the request is complete, dispose them all immediately.
21:25:41  <trevnorris>s/request/write
21:26:05  <trevnorris>then they can be tracked internally, and you don't have to loop through a js array.
21:27:33  <MI6>nodejs-master-windows: #5 UNSTABLE windows-x64 (9/587) windows-ia32 (10/587) http://jenkins.nodejs.org/job/nodejs-master-windows/5/
21:31:15  <indutny>trevnorris: probably a good idea
21:31:20  <indutny>but sounds complicated
21:31:27  <indutny>still might be worth trying it
21:31:40  <trevnorris>i'll give it a quick go. if it gets too complicated then i'll drop it.
21:31:53  <trevnorris>just wanted to make sure you weren't opposed to it. :)
21:34:44  * c4miloquit (Remote host closed the connection)
21:35:48  <indutny>isaacs: a question :)
21:35:48  <indutny>hi
21:35:50  <indutny>https://github.com/isaacs/npm/issues/2063
21:36:03  <indutny>is it still an issue? (circular npm dependencies)
21:36:18  <indutny>if yes - will you consider reviewing pull request that'll fix it?
21:38:06  * rendarquit
21:41:19  * kenperkinsquit (Quit: Textual IRC Client: http://www.textualapp.com/)
21:44:03  * hzquit
21:45:24  <MI6>joyent/node: Trevor Norris master * b7f6e6b : stream_wrap: remove unused arg from WriteBuffer - http://git.io/h4Lcqg
21:50:31  <MI6>joyent/libuv: Ben Noordhuis master * 4a8ec0c : sunos: handle disabled watcher after callback - http://git.io/hw-FxA
21:51:17  <trevnorris>isaacs: dude, finally got permission to send out the email. it's a miracle!!!
21:55:06  <bnoordhuis>trevnorris: don't keep me in suspense, what email?
21:55:29  <trevnorris>bnoordhuis: that i'm officially a mozilla sponsored contributor.
21:55:40  <bnoordhuis>ah, congrats :)
21:55:52  <trevnorris>thanks. :)
21:55:58  <TooTallNate>trevnorris: so you're node full time now?
21:56:03  <trevnorris>TooTallNate: yup.
21:56:08  <TooTallNate>awesome :)
21:56:12  <TooTallNate>congrats
21:56:17  <trevnorris>actually have been for a couple months. it's just taken a long time to get the hr stuff done.
21:56:45  <MI6>libuv-master-gyp: #26 UNSTABLE windows-x64 (6/189) smartos-x64 (3/188) smartos-ia32 (3/188) linux-x64 (1/188) windows-ia32 (4/189) linux-ia32 (1/188) http://jenkins.nodejs.org/job/libuv-master-gyp/26/
21:57:12  <MI6>libuv-master: #86 UNSTABLE osx (1/189) linux (1/188) windows (4/189) smartos (3/188) http://jenkins.nodejs.org/job/libuv-master/86/
21:57:14  <trevnorris>\o/
21:57:34  <trevnorris>being that my manager's manager left, then the CEO resigned. took a while to get this through.
21:58:07  <amartens>HR paperwork can be slow at the best of times
21:58:08  <bnoordhuis>the ceo resigned over you working on node? that's heavy
21:58:20  <trevnorris>lol
21:58:36  * `3rdEdenquit (Remote host closed the connection)
21:58:51  <trevnorris>he resigned because B2G successfully launched. wanted to move on to something else.
21:59:13  <tjfontaine>"oh god, my shit is in production, gotta leave before they ask me to support it"
21:59:30  <tjfontaine>that's basically how I feel about my software as well
21:59:35  * kenperkinsjoined
21:59:38  <trevnorris>heh
22:00:43  <trevnorris>TooTallNate: tell you the truth, been working on node almost full time since Jan.
22:00:54  <TooTallNate>trevnorris: hehe :)
22:01:06  <trevnorris>my last project finished up, so started on that buffer float write patch. then just kept going with it.
22:01:22  <TooTallNate>i remember :)
22:02:58  * benoitcquit (Read error: Operation timed out)
22:03:36  * piscisaureus_quit (Ping timeout: 256 seconds)
22:06:11  <MI6>joyent/node: Brian White master * aca275f : stream_wrap: fix memory leak on early return - http://git.io/5PWXcw
22:06:41  * benoitcjoined
22:10:08  <tjfontaine>damn you python, damn you so very much
22:15:00  <MI6>nodejs-master: #199 UNSTABLE osx-ia32 (1/587) http://jenkins.nodejs.org/job/nodejs-master/199/
22:16:05  * pachetquit (Quit: leaving)
22:17:36  * kazuponjoined
22:19:19  <tjfontaine>bnoordhuis: sigh, lgty? https://github.com/tjfontaine/node/compare/test-dur-master
22:20:19  <bnoordhuis>tjfontaine: almost. long line and `from foo` lines should be grouped
22:21:10  <tjfontaine>bnoordhuis: do you want me to group with the from at the top or the from at the bottom, or move them altogether? :)
22:21:23  <bnoordhuis>tjfontaine: move the new one near the old one
22:21:32  <tjfontaine>there are two others
22:21:36  <tjfontaine>from os.path import join, dirname, abspath, basename, isdir, exists
22:21:40  <tjfontaine>isn't in context of the diff
22:21:49  <bnoordhuis>i think google's python styleguide mandates that normal imports go first, followed by `from x` imports
22:21:57  * kazuponquit (Ping timeout: 248 seconds)
22:22:05  <tjfontaine>yes it does, I can move them all
22:22:21  <bnoordhuis>that' be good
22:22:27  <bnoordhuis>*d
22:22:30  * bnoordhuissighs
22:23:15  <tjfontaine>refresh'd https://github.com/tjfontaine/node/compare/test-dur-master
22:23:20  <tjfontaine>oh long line
22:23:53  <tjfontaine>is there a style on new lines and binary operators?
22:24:14  <tjfontaine>I'll jsut make them multiple statements
22:28:50  <bnoordhuis>tjfontaine: oh, in python you can just do x = (bla * bla * <newline><indent> bla * bla)
22:29:07  <tjfontaine>ya I wasn't sure if that's clean though
22:29:39  <bnoordhuis>mwah, as long as it's readable
22:30:12  <tjfontaine>also was using total_seconds wrong as it turns out, because its name lies as well
22:31:21  <tjfontaine>ok +pushed https://github.com/tjfontaine/node/compare/test-dur-master
22:32:40  <trevnorris>anyone know why Readable() in _stream_readable does a Stream.call(this) instead of EventEmitter.call(this)?
22:32:41  <bnoordhuis>tjfontaine: lgtm
22:32:54  <tjfontaine>bnoordhuis: thanks
22:33:00  <trevnorris>one just calls the other, and it adds an extra .call() in there.
22:34:33  <MI6>joyent/node: Timothy J Fontaine master * 56492de : test: don't use total_seconds() because of py2.6 - http://git.io/p9re5g
22:36:02  * piscisaureus_joined
22:42:48  <bnoordhuis>piscisaureus_: those cluster-roud-robin test failures also happen in master, it seems
22:44:12  <tjfontaine>https://github.com/tjfontaine/node/compare/joyent:v0.10...t-s-10 not enough contexet in that diff, but because of the `import utils` coming from a different path it's grouped differently in v0.10, are you ok with that bnoordhuis?
22:44:34  <MI6>nodejs-master-windows: #6 UNSTABLE windows-x64 (9/587) windows-ia32 (10/587) http://jenkins.nodejs.org/job/nodejs-master-windows/6/
22:46:13  <bnoordhuis>tjfontaine: oh, sure
22:46:22  <bnoordhuis>shouldn't this have gone first into v0.10 though?
22:46:44  <bnoordhuis>i.e. v0.10, then merge into master
22:46:58  <tjfontaine>oh, I have been doing cherry-picks
22:47:16  <tjfontaine>but I can change my workflow
22:50:29  <tjfontaine>anyway I've already broken it for this round, so I'll change going forward
22:50:35  <MI6>joyent/node: Timothy J Fontaine v0.10 * cf87ee6 : test: don't use total_seconds() because of py2.6 - http://git.io/D9nP-w
22:56:23  * loladiroquit (Quit: loladiro)
22:57:54  <MI6>nodejs-master: #200 UNSTABLE smartos-ia32 (3/587) smartos-x64 (3/587) http://jenkins.nodejs.org/job/nodejs-master/200/
22:58:35  * c4milojoined
23:00:48  * AvianFluquit (Read error: Connection reset by peer)
23:01:25  * AvianFlujoined
23:03:20  <MI6>libuv-node-integration: #44 UNSTABLE osx-x64 (2/587) osx-ia32 (3/587) linux-ia32 (2/587) windows-ia32 (8/587) windows-x64 (8/587) http://jenkins.nodejs.org/job/libuv-node-integration/44/
23:18:11  * kazuponjoined
23:18:26  <trevnorris>isaacs: where the hell are you man? mozilla will be releasing a blog post soon about me working on node.
23:18:40  <isaacs>trevnorris: yo
23:18:47  <isaacs>trevnorris: awesome
23:18:52  <TooTallNate>double-post to blog.nodejs.org?
23:19:06  <trevnorris>isaacs: send off the email at your pleasure. everything is good on this end. :)
23:19:20  <isaacs>TooTallNate: we don't usually have blog posts about new core committers, just email to the mailing list.
23:19:29  <TooTallNate>isaacs: good point
23:20:43  <tjfontaine>isaacs: fwiw you may need to `make clean` your crypto pr isn't building atm
23:21:13  <isaacs>orly?
23:21:18  <isaacs>weird..
23:21:19  <tjfontaine>http://jenkins.nodejs.org/job/node-pullrequest/DESTCPU=x64,label=osx/315/console
23:21:33  <tjfontaine>error: no viable conversion from 'Handle<v8::Object>' to 'Local<v8::Object>'
23:21:49  <tjfontaine>../src/string_bytes.cc:185:19
23:22:55  * kazuponquit (Ping timeout: 260 seconds)
23:26:47  <isaacs>d'oh
23:26:54  <isaacs>how do you convert from Handle to local?
23:27:18  <isaacs>or maybe i should just revert that change
23:27:19  <trevnorris>Local<Object>::New(item, node_isolate);
23:27:23  <isaacs>kk
23:27:34  <trevnorris>isaacs: are you setting up arguments to pass to js?
23:27:48  <isaacs>trevnorris: Local<Object> valObj = Local<Object>::New(val.As<Object>(), node_isolate);
23:28:01  <isaacs>trevnorris: no, i'm creating a single thing rather than doing val.As<Object>() twice
23:28:14  <isaacs>trevnorris: since I need the Buffer::Data() and also the Buffer::Length()
23:28:26  <trevnorris>ah, ok.
23:29:03  <isaacs>hrm...
23:29:04  <isaacs>../src/string_bytes.cc:185: error: no matching function for call to ‘v8::Local<v8::Object>::New(v8::Handle<v8::Object>, v8::Isolate*&)’
23:29:07  <isaacs>../deps/v8/include/v8.h:4197: note: candidates are: static v8::Local<T> v8::Local<T>::New(v8::Handle<T>) [with T = v8::Object]
23:29:10  <isaacs>looks like itdoesn't like getting the isolate?
23:29:10  <trevnorris>oh, duh
23:29:16  <trevnorris>switch the isolate...
23:29:58  <trevnorris>remembering when isolate comes first and last is a PITA
23:30:15  <tjfontaine>it is a bit of a silly situation
23:31:10  <isaacs>tjfontaine: seems like it just doesn't like gtting an isolate at all. only option is v8::Local<T>::New(v8::Handle<T>)
23:31:21  <isaacs>er, trevnorris ^
23:31:27  <tjfontaine>yes I don't think the isolate is needed in this case
23:31:30  * isaacsshakes fist at same-length t-names
23:31:39  <trevnorris>oh fuck. no, I forget this is for the v0.10 branch
23:31:41  <tjfontaine>but I have an @! :P
23:31:45  <trevnorris>so you don't need it at all
23:31:52  <isaacs>yep
23:32:13  <isaacs>trevnorris: that reminds me, is your handle registered in irc?
23:32:30  <trevnorris>isaacs: yup
23:32:47  <isaacs>there we go
23:42:22  <MI6>nodejs-master-windows: #7 UNSTABLE windows-x64 (9/587) windows-ia32 (10/587) http://jenkins.nodejs.org/job/nodejs-master-windows/7/
23:45:35  <isaacs>trevnorris: it is official :)
23:45:48  <trevnorris>whoot!!! feels good to have that done.
23:45:59  <trevnorris>and the internal email was sent out about 10 mins ago.
23:46:23  <isaacs>oh, and i guess i misspoke, you started out doing buffer/typedarray mumbo jumbo, right?
23:46:29  <isaacs>not the node.cc/node.js stuff
23:47:16  <trevnorris>eh, technically.
23:47:58  <TooTallNate>trevnorris: ya, you're the one who told me to use ArrayBuffer and Buffer interchangeably :p
23:48:02  <TooTallNate>now that's screwed me
23:48:09  <trevnorris>....
23:48:13  * trevnorrishides in a corner
23:48:15  <TooTallNate>hahaha
23:48:16  <TooTallNate>:p
23:48:17  <TooTallNate>it's ok
23:48:36  <TooTallNate>i just gotta use Buffer / Uint8Array interchangeably
23:48:43  <TooTallNate>which kinda sucks since I can't create "views" on top
23:48:45  * bajtosquit (Quit: bajtos)
23:48:45  <TooTallNate>but meh
23:48:51  <TooTallNate>i'm not gonna get my way :p
23:49:06  <trevnorris>well, it's not that it's impossible.
23:49:13  <TooTallNate>or just use ArrayBuffers religiously
23:49:37  <TooTallNate>trevnorris: sure, i mean we could make ArrayBuffer's on top of Buffer's, no?
23:50:04  <trevnorris>TooTallNate: oh, I don't think you've heard this discussion.
23:50:18  <TooTallNate>what'd i miss?
23:50:28  <trevnorris>TooTallNate: um.... v8 is starting to support typed arrays natively. and when that happens there'll be _no_ interoperability between the two
23:50:40  <trevnorris>unless we create a Buffer.toArrayBuffer() whatnot
23:50:53  * bnoordhuisquit (Ping timeout: 245 seconds)
23:50:55  <TooTallNate>oh right, ya i knew about that… i guess that would kill the ArrayBuffer inherits from Buffer idea
23:51:11  <tjfontaine>well, not necessarily
23:51:28  <tjfontaine>there's always the chance that we convince them to use Buffer as the backing store for ArrayBuffers :)
23:51:47  <trevnorris>tjfontaine: but v8 is supporting them natively, so we can't control that.
23:52:07  * piscisaureus_quit (Ping timeout: 245 seconds)
23:52:09  <tjfontaine>trevnorris: we could try and participate, getting the "native" buffer type that piscisaureus_ and indutny were looking into
23:52:12  <tjfontaine>etc
23:52:31  <MI6>nodejs-v0.10: #176 UNSTABLE smartos-x64 (4/582) http://jenkins.nodejs.org/job/nodejs-v0.10/176/
23:52:47  <trevnorris>can give it a shot. they're array buffer implementation is slow as shiz
23:53:05  <tjfontaine>mostly because it's required to memset/memcpy?
23:53:26  <trevnorris>also because they set a bunch of read only properties internally.
23:53:36  <trevnorris>those really kill performance right now.
23:53:46  <trevnorris>a bunch as in 3
23:53:58  <tjfontaine>right, but anyway this would be behind that world
23:54:06  <trevnorris>tjfontaine: also, we'd have to change Buffer .slice to make a copy instead of pointing.
23:54:20  <trevnorris>oh, yeah.
23:54:32  <tjfontaine>trevnorris: not necessarily, when it gets promoted to arraybuffer it just changes the prototype for it
23:54:46  <trevnorris>true true
23:55:07  <trevnorris>that's actually something similar to my allocator now.
23:55:12  <tjfontaine>yes
23:57:40  * perezdjoined