00:12:08  * lwillejoined
00:17:58  <lwille>hi there! Is there a way I can ensure that async workers invoked by uv_queue_work are called serially? Some kind of mutex comes to my mind ... the async workers are accessing hardware which can't be done in parallel, so I have to block other threads while one worker is accessing the device.
00:20:35  <ryah_>lwille: we dont have a way to do that currently
00:20:46  <ryah_>you'd need to dispatch them one after the previous completes
00:22:43  * `3rdEdenchanged nick to `3E|Shower
00:26:20  <lwille>just used pthread_mutex_lock and pthread_mutex_unlock - just does the job.
00:27:11  <lwille>well, I didn't want to exec them in series, but just avoid concurrent hardware access, so I wrapped the hardware calls with mutex_lock and mutex_unlock.
00:31:13  <ryah_>it'd be nicer if we had a knob for determining the thread pool size
00:31:19  <ryah_>and you could just set that to 1
00:34:05  * lwillequit (Read error: Connection reset by peer)
00:34:12  <piscisaureus__>we would need to have custom thread pools first
00:34:25  * lwillejoined
00:37:38  <lwille>indeed, but I think that mutexes are the best for my situation. Or is there any advice not to use them with uv?
00:38:28  * `3E|Showerchanged nick to `3rdEden
00:40:18  <piscisaureus__>lwille: no it's fine to use them
00:40:40  <piscisaureus__>lwille: although you should use uv_mutex for x-platform
00:40:42  <piscisaureus__>:-)
00:46:49  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
00:48:05  * piscisaureus__quit (Quit: ~ Trillian Astra - www.trillian.im ~)
00:50:00  * lwillequit (Read error: Connection reset by peer)
00:50:04  * leonhardtwillejoined
00:52:03  <leonhardtwille>piscisaureus__ , I know that it's not the nicest thing to do :) But what could I do instead? I think it would be worse to retry every failed call to the driver (because concurrent calls tend to fail) instead of locking the resource.
00:56:15  <bnoordhuis>leonhardtwille: mutexes are fine if you want linearity
00:56:33  <bnoordhuis>just don't block too long in your critical sections
01:07:50  <leonhardtwille>is there any reason why uv_mutex is not included in 0.6.8 ?
01:08:07  * leonhardtwillechanged nick to lwille
01:12:42  * `3rdEdenquit (Quit: Leaving...)
01:22:21  * `3rdEdenjoined
01:34:36  <ryah_>lwille: we don't change API in the v0.6 branch
01:34:45  <ryah_>and it wasn't there when it was branched
01:35:01  <ryah_>it's part of the work to support isolates
01:42:13  <lwille>okay, thanks :)
01:42:22  * lwillequit (Quit: Leaving...)
01:42:35  * lwillejoined
01:42:39  * lwillequit (Client Quit)
01:48:31  * `3rdEdenquit (Quit: Zzzzzz)
01:52:55  <CIA-115>libuv: Igor Zinkovsky v0.6 * r726e36c / src/win/udp.c : implement uv_udp_set_multicast_ttl and uv_udp_set_broadcast on windows - http://git.io/Vtcyfg
01:54:36  * travis-cijoined
01:54:37  <travis-ci>[travis-ci] joyent/libuv#45 (v0.6 - 726e36c : Igor Zinkovsky): The build is still failing.
01:54:37  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/497b1ec...726e36c
01:54:37  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/560328
01:54:37  * travis-cipart
01:56:33  * orlandovftwjoined
02:00:01  * brsonquit (Quit: leaving)
02:00:51  * bnoordhuisquit (Ping timeout: 255 seconds)
02:00:55  <CIA-115>libuv: Igor Zinkovsky v0.6 * redbabe6 / src/win/udp.c : windows: set error using WSAGetLastError - http://git.io/Z_0hKA
02:02:39  * travis-cijoined
02:02:39  <travis-ci>[travis-ci] joyent/libuv#46 (v0.6 - edbabe6 : Igor Zinkovsky): The build is still failing.
02:02:39  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/726e36c...edbabe6
02:02:39  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/560344
02:02:39  * travis-cipart
02:09:20  * orlandovftwquit (Ping timeout: 252 seconds)
02:21:50  * mmaleckichanged nick to mmalecki[zzz]
02:48:48  * mikealquit (Quit: Leaving.)
02:49:18  * mikealjoined
03:38:17  * sh1mmerjoined
05:26:38  * sh1mmerquit (Quit: sh1mmer)
05:27:19  * sh1mmerjoined
05:29:36  * isaacsquit (Remote host closed the connection)
06:02:55  * mralephjoined
06:19:27  * orlandovftwjoined
06:29:06  * orlandovftwquit (Quit: leaving)
06:29:21  * orlandovftwjoined
06:43:33  * orlandovftwquit (Quit: leaving)
06:43:46  * orlandovftwjoined
06:45:43  * isaacsjoined
06:53:59  * mralephquit (Quit: Leaving.)
07:12:02  <indutny>isaacs: hi! :)
07:12:15  <isaacs>indutny: ello
07:13:36  <indutny>isaacs: i think zlib is still leaking
07:13:39  <indutny>isaacs: https://github.com/joyent/node/issues/2595
07:14:00  <indutny>isaacs: I'll modify example code
07:14:02  <indutny>one sec
07:14:20  <indutny>ah
07:14:24  <indutny>wrong read your message
07:14:31  <indutny>sorry
07:15:36  <isaacs>yeah, i'm gonna merge 0.6 into master tomorrow ("tonight" for you, i guess)
07:16:12  <isaacs>need to push out 0.7.1
07:16:14  <indutny>isaacs: I already merged it :)
07:16:21  <isaacs>oh, just this fix, or everything?
07:16:23  <indutny>isaacs: but I need uv guys to merg uv first
07:16:30  <indutny>isaacs: everything
07:16:41  <isaacs>ok, awesome :)
07:16:44  * mmalecki[zzz]changed nick to mmalecki
07:16:45  <indutny>isaacs: git merge -X patience v0.6
07:16:47  <indutny>:)
07:16:52  <isaacs>what's -X patience?
07:17:10  <indutny>With this option, merge-recursive spends a little extra time to avoid mismerges that sometimes occur due to unimportant matching lines (e.g., braces from distinct functions). Use this when the branches to be merged have diverged wildly. See also git-diff(1) --patience.
07:17:17  <indutny>cool thing
07:17:21  <isaacs>oh, nice
07:17:40  <isaacs>i usually just get the conflicts, and painstakingly learn about each change
07:17:42  <isaacs>which is good too.
07:17:59  <isaacs>sometimes it takes a big merge conflict to get you to understand the program you're maintaining :)
07:18:23  <indutny>isaacs: yeah :)
07:18:43  <indutny>isaacs: I backported changes from bleeding_edge to v8's 3.6 branch
07:18:48  <indutny>isaacs: a BIG BIG merge
07:19:07  <indutny>isaacs: and a lot of conflicts
07:19:20  <indutny>isaacs: I'm just running git diff origin/master before commititng
07:19:29  <indutny>committing*
07:19:57  <isaacs>yikes
07:28:01  * kuebkjoined
07:38:16  * harry__joined
07:41:10  * sj26joined
07:41:23  * kuebk1joined
07:43:33  * kuebkquit (Ping timeout: 248 seconds)
08:45:35  <harry__>isaacs, are you on this channel?
08:46:39  <isaacs>harry__: yes
08:48:19  <harry__>I'm the edgit on node.js talking about tutorials etc.
09:35:05  <CIA-115>node: Igor Zinkovsky v0.6 * rf5e742b / deps/uv/src/win/udp.c : upgrade libuv to edbabe6f83 - http://git.io/Cdt9NA
09:42:25  * paddybyersjoined
09:42:45  * travis-cijoined
09:42:45  <travis-ci>[travis-ci] joyent/node#301 (v0.6 - f5e742b : Igor Zinkovsky): The build passed.
09:42:45  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/9edb984...f5e742b
09:42:45  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/561693
09:42:45  * travis-cipart
09:43:05  <CIA-115>node: Igor Zinkovsky v0.6 * r6c0c00a / src/node.h : fix windows build - http://git.io/hIzH6g
09:48:27  * dshaw_quit (Quit: Leaving.)
09:50:46  * travis-cijoined
09:50:47  <travis-ci>[travis-ci] joyent/node#302 (v0.6 - 6c0c00a : Igor Zinkovsky): The build passed.
09:50:47  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/f5e742b...6c0c00a
09:50:47  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/561721
09:50:47  * travis-cipart
09:54:54  * orlandovftwquit (Ping timeout: 240 seconds)
10:00:17  * lwillejoined
10:21:34  * isaacsquit (Remote host closed the connection)
11:50:51  * paddybyers_joined
11:51:14  * paddybyersquit (Ping timeout: 272 seconds)
11:51:14  * paddybyers_changed nick to paddybyers
12:00:31  * sj26quit (Excess Flood)
12:04:08  * sj26joined
12:14:25  * piscisaureus_joined
12:22:27  * bnoordhuisjoined
12:23:35  <indutny>bnoordhuis: hi!
12:23:46  <indutny>bnoordhuis: can you please merge uv v0.6 into uv's master
12:24:33  <bnoordhuis>indutny: i suppose
12:24:49  <bnoordhuis>did you merge node v0.6 and master with -s ours?
12:25:15  <indutny>bnoordhuis: nope, with -X patience
12:25:36  <indutny>bnoordhuis: I merge was ... strange
12:25:41  <indutny>bnoordhuis: with ours ^
12:26:10  <bnoordhuis>yeah, i always resolve merge conflicts manually
12:26:30  <bnoordhuis>doing it automatically nearly always does the wrong thing
12:26:54  <indutny>yeah
12:26:59  <indutny>sometimes it's useful
12:27:30  <bnoordhuis>indutny: where there merge conflicts under deps?
12:27:37  <indutny>bnoordhuis: yes
12:27:45  <indutny>bnoordhuis: I just chosen master's one
12:27:51  <indutny>bnoordhuis: for all deps files
12:27:52  <bnoordhuis>good
12:28:11  <indutny>bnoordhuis: I'll show you commit first, before pushing to master anyway ;)
12:28:51  <CIA-115>libuv: Dan VerWeire master * rb674187 / src/unix/udp.c : unix: set SO_REUSEADDR before binding - http://git.io/i2psSg
12:28:51  <CIA-115>libuv: Dan VerWeire master * r497b1ec / (include/uv.h src/unix/udp.c): unix: add uv_udp_set_broadcast() and uv_udp_set_multicast_ttl() - http://git.io/ccCkQw
12:28:51  <CIA-115>libuv: Igor Zinkovsky master * r726e36c / src/win/udp.c : implement uv_udp_set_multicast_ttl and uv_udp_set_broadcast on windows - http://git.io/Vtcyfg
12:28:52  <CIA-115>libuv: Igor Zinkovsky master * redbabe6 / src/win/udp.c : windows: set error using WSAGetLastError - http://git.io/Z_0hKA
12:28:52  <CIA-115>libuv: Ben Noordhuis master * r1cca230 / (include/uv.h src/unix/udp.c src/win/udp.c): Merge remote-tracking branch 'origin/v0.6' - http://git.io/ZW_HNQ
12:29:22  <indutny>cool!
12:30:42  * travis-cijoined
12:30:42  <travis-ci>[travis-ci] joyent/libuv#47 (master - 1cca230 : Ben Noordhuis): The build is still failing.
12:30:42  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/517bfc8...1cca230
12:30:42  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/562370
12:30:42  * travis-cipart
12:31:53  <indutny>bnoordhuis: https://github.com/indutny/node/commit/c31162a39f871e26b83eb7adb865f62d520f9aeb
12:32:47  <bnoordhuis>indutny: there shouldn't be changes under deps/uv
12:32:56  <bnoordhuis>upgrade deps/uv first, then merge
12:32:57  <indutny>bnoordhuis: ohhhh
12:33:01  <indutny>bnoordhuis: really?
12:33:07  <bnoordhuis>yes
12:33:15  <indutny>bnoordhuis: I just merged master in same commit
12:33:21  <indutny>fck
12:33:36  <bnoordhuis>want me to update deps/uv?
12:33:36  <indutny>ok, I'll come back in 20 minutes
12:33:49  <indutny>bnoordhuis: and rebase then?
12:33:58  <bnoordhuis>yeah
12:34:01  <indutny>bnoordhuis: sounds good
12:34:08  <indutny>bnoordhuis: will rebase work fine with merge?
12:34:17  <indutny>bnoordhuis: I think merge commit will have one parent after rebase
12:34:39  <bnoordhuis>indutny: just nuke and redo your merge
12:34:48  <indutny>bnoordhuis: nuke?
12:34:54  <bnoordhuis>indutny: undo :)
12:35:11  <indutny>bnoordhuis: heh
12:35:32  <indutny>bnoordhuis: ok, 20-30 minutes :D
12:35:37  <indutny>bnoordhuis: I've some stuff to do before it
12:36:04  <CIA-115>node: Ben Noordhuis master * r2433eeb / (22 files in 7 dirs): uv: upgrade to 1cca230 - http://git.io/CyegSA
12:36:06  <bnoordhuis>sure, no rush
12:46:49  <indutny>bnoordhuis: 120000 insertions / sec with bulk_set in bplus ;)
12:47:55  <indutny>bnoordhuis: do you have time to take a short look at the code?
12:48:06  <bnoordhuis>indutny: well, not really
12:48:18  <indutny>bnoordhuis: np
12:48:21  <indutny>just asking
12:49:51  * travis-cijoined
12:49:51  <travis-ci>[travis-ci] joyent/node#303 (master - 2433eeb : Ben Noordhuis): The build is still failing.
12:49:51  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/ff51263...2433eeb
12:49:51  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/562408
12:49:51  * travis-cipart
13:31:55  * sj26quit (Excess Flood)
13:35:40  * sj26joined
14:54:08  * mikeal1joined
14:54:09  * mikealquit (Read error: Connection reset by peer)
15:09:10  * AndreasMadsenjoined
15:29:28  * sh1mmerquit (Quit: sh1mmer)
15:32:24  * kuebk1part
15:42:22  <AndreasMadsen>So it takes 3 sec to write a 20 MB message from fork IPC, and the process.send is sync - hmm, that blocks :/
15:44:59  <AndreasMadsen>https://gist.github.com/1663885
15:59:10  * mikeal1quit (Read error: Connection reset by peer)
15:59:18  * mikealjoined
16:05:15  * `3rdEdenjoined
16:14:17  * mikeal1joined
16:14:17  * mikealquit (Read error: Connection reset by peer)
16:20:19  * pieternjoined
16:24:10  * mikealjoined
16:24:10  * mikeal1quit (Read error: Connection reset by peer)
16:24:58  * sh1mmerjoined
16:29:29  <AndreasMadsen>^-- https://github.com/joyent/node/issues/2598
16:33:32  * mrb_bkquit (Read error: Connection reset by peer)
16:33:32  * Raynosquit (Read error: Connection reset by peer)
16:36:46  * mikeal1joined
16:36:49  * mikealquit (Read error: Connection reset by peer)
16:39:55  * isaacsjoined
16:40:40  * mrb_bkjoined
16:43:29  <isaacs>good morning
16:43:39  <bnoordhuis>good afternoon
16:43:54  <piscisaureus_>isaacs: good afternoon
16:44:06  <piscisaureus_>isaacs: are you not supposed to be on stage or something right now?
16:44:15  <mmalecki>I never know when does afternoon start :<
16:44:22  <isaacs>nono, that's evening pacific time
16:44:28  <mmalecki>so, good (.+)
16:44:34  <bnoordhuis>mmalecki: i don't know about poland but noon starts at 1200 hours here
16:44:50  <isaacs>like 12 hours
16:45:01  <piscisaureus_>in poland noon starts 5 hours befoore they start drinking
16:45:23  <bnoordhuis>piscisaureus_: more like two hours after
16:45:24  <piscisaureus_>:-p
16:45:26  <mmalecki>bnoordhuis: but saying 'afternoon' right after 1200 doesn't feel right :<
16:45:41  <mmalecki>well, first thing worth mentioning.
16:45:46  <mmalecki>we never actually stop drinking
16:46:24  <isaacs>so, between now and nodeup live, my goal is to finish up the doc/website reskin and get that live, and release 0.7.1.
16:46:53  <mmalecki>would be cool to get my libuv patch in
16:46:57  <piscisaureus_>isaacs: website gets another reskin?
16:47:12  <isaacs>piscisaureus_: the frontpage isn't change much
16:47:18  <isaacs>but the docs and stuff are much nicer.
16:47:25  <mmalecki>https://github.com/joyent/libuv/pull/295 <- this one (need some windows assistance)
16:47:37  <piscisaureus_>isaacs: http://nodemanual.org/latest/nodejs_ref_guide/assert.html
16:47:39  <bnoordhuis>that's your cue, piscisaureus_
16:48:37  <isaacs>neat :)
16:48:57  <piscisaureus_>isaacs: I think it is not officially released yet
16:49:06  <isaacs>kewl
16:49:19  <piscisaureus_>oh I have standup bbiab
16:50:24  <isaacs>still a few pixels to push around, but here's the general idea:
16:50:25  <isaacs>http://static.izs.me/nodejs.org-reskin-2012/api/process.html
16:50:34  <mmalecki>bnoordhuis: oh, btw, this fork stuff. we actually went to use your idea :)
16:50:41  <isaacs>http://static.izs.me/nodejs.org-reskin-2012/community/index.html
16:50:47  <mmalecki>bnoordhuis: we wrote our little daemonizer: https://github.com/AvianFlu/aeternum
16:51:25  <mmalecki>hah, NIH <3
16:52:36  <isaacs>http://static.izs.me/nodejs.org-reskin-2012/about/index.html
16:53:36  <raggi>heh, i wrote one of those, although i did the double fork dance
16:53:56  <isaacs>piscisaureus_: i like separating out modules/globals/concepts
16:54:12  <isaacs>piscisaureus_: someone was suggesting that just the other day, actually. it's grown super messy, what we ahve now.
16:54:29  * Raynosjoined
16:55:49  <bnoordhuis>mmalecki: very good :)
16:56:41  <bnoordhuis>oh, service announcement: i'll be good for most of the day tomorrow and the day after on account of a funeral
16:56:46  <bnoordhuis>s/good/gone/
16:57:07  <bnoordhuis>i'll be good too
16:57:07  <AndreasMadsen>indutny: I just updated the disconnect patch -> https://github.com/joyent/node/pull/2591
16:57:21  <mmalecki>bnoordhuis: not sure when will forever use it tho
17:00:12  <piscisaureus_>isaacs: yeah
17:00:23  <AndreasMadsen>so lets talk process.send - why is it sync?
17:00:25  <piscisaureus_>isaacs: sorry was afk. The docs needs a refactor
17:00:31  <isaacs>piscisaureus_: yeah
17:00:48  <isaacs>i think the content changes especially that you have here should be pulled into the main node repo
17:01:00  <piscisaureus_>isaacs: yes we are planning on doing that
17:01:06  <isaacs>kewl
17:01:11  <isaacs>i have strong feelings about comment.
17:01:14  <isaacs>comments
17:01:26  <piscisaureus_>isaacs: but they are busy finishing that website atm (it's not public yet)
17:01:41  <piscisaureus_>isaacs: the comments do not need to be pulled in
17:02:02  <piscisaureus_>isaacs: if you want to talk to the technical writer who did most of the editing I can hook you guys up
17:02:14  <isaacs>cool
17:02:34  <isaacs>the tricky thing about large-scale changes to docs is that you really need someone close to the project to either do it, or read every word and check it.
17:02:41  <isaacs>i've gotten bit by that in the past, at yahoo.
17:02:54  <piscisaureus_>isaacs: yeah, it will have to be a pull request and a review
17:02:58  <isaacs>yep
17:03:25  <isaacs>another thought that occurs to me, having been involved in this reskin... it's WAY too hard to do right now.
17:03:27  <piscisaureus_>isaacs: right now the stuff is still under "internal review" e.g. some experienced node users are reviewing it before nodemanual goes live
17:03:31  <isaacs>our html stuff is a mess in core.
17:04:04  <mmalecki>piscisaureus_: hey, mind taking a look at https://github.com/joyent/libuv/pull/295 ?
17:04:08  <isaacs>i'd like to get to a single CSS file and set of images that can be used by the blog, api docs, and website, and potentially also use it for npm modules.
17:05:00  <isaacs>but right now it's just a huge mess of hacks.
17:05:29  <mmalecki>I think you just described CSS as a language :)
17:05:33  <isaacs>matt has done a great job of wading in and getting it most of the way there, but there's always like a million little edge cases to watch out for, because it's had so much surgery
17:07:13  <piscisaureus_>isaacs: I'd like to get the content changes we have so far reviewed first
17:07:18  <piscisaureus_>isaacs: do restructuring then
17:07:31  <piscisaureus_>because otherwise it will be impossible to rebase that stuff
17:07:53  <isaacs>yeah
17:07:59  <isaacs>well, restructuring doesn't touch the markdown content.
17:08:16  <isaacs>but yeah, if you guys are messing with the doctool, then that'll collide
17:08:17  <piscisaureus_>oh, right
17:08:20  <piscisaureus_>we are not
17:08:26  <isaacs>it's mostly just the template and css files.
17:08:32  <piscisaureus_>isaacs: the site is generated straight of the markdown I think
17:09:02  <isaacs>the troublesome thing is that doc/api_assets/style.css and doc/pipe.css are *almost* the same, except for like a pixel here and a pixel there, so any changes are not portable between them
17:09:37  <piscisaureus_>heh
17:09:46  <piscisaureus_>yes, nobody cares about css
17:09:49  <piscisaureus_>at least not ryah
17:15:37  <mmalecki>piscisaureus_: hey, wondering about windows implementation of https://github.com/joyent/libuv/pull/295, mind taking a look?
17:16:43  <mmalecki>piscisaureus_: I know there's ERROR_BUFFER_OVERFLOW, but not sure if can it happen without file name being too long
17:16:52  <piscisaureus_>mmalecki: yes it can happen in many cases
17:17:18  <piscisaureus_>let me thing
17:17:53  * mikeal1quit (Quit: Leaving.)
17:18:39  <piscisaureus_>mmalecki: yes it can also mean ENOMEM or EFAULT in some cases
17:19:19  <mmalecki>piscisaureus_: I see. so is there a way to reliably implement ENAMETOOLONG on windows?
17:22:14  <indutny>what's happening with github?
17:22:26  <mmalecki>indutny: pages are down
17:22:30  <mmalecki>http://status.github.com/
17:22:42  <indutny>mmalecki: I can't load status :D
17:22:45  <piscisaureus_>mmalecki: well add a specific mapping after the CreateFile call
17:22:48  <indutny>mmalecki: and ssh isn't working
17:22:58  <piscisaureus_>mmalecki: as well as any other calls that take a filename
17:23:59  * paddybyers_joined
17:23:59  * paddybyersquit (Read error: Connection reset by peer)
17:24:00  * paddybyers_changed nick to paddybyers
17:24:04  <mmalecki>piscisaureus_: hm, I see. looks like a lot of effort?
17:24:21  <piscisaureus_>mmalecki: it depends what you consider a lot of effort
17:24:34  <piscisaureus_>mmalecki: it's like adding an if ( in 5 different places or so
17:25:17  <mmalecki>piscisaureus_: hm, I see. could you possibly write a patch for that? I don't have any machine I could test this stuff on.
17:25:30  <mmalecki>piscisaureus_: also, would it be possible to merge Unix patch now?
17:26:11  <mmalecki>indutny: ssh is back up for me
17:26:29  <mmalecki>they need a status page for their balancers as well ;)
17:28:31  <indutny>mmalecki: yeah
17:28:44  <piscisaureus_>hmm I am busy
17:28:54  <piscisaureus_>indutny: can you land mmalecki's ENAMETOOLONG patch?
17:29:10  <indutny>piscisaureus_: of course, where is it?
17:29:15  <piscisaureus_>mmalecki ^
17:29:19  <mmalecki>https://github.com/joyent/libuv/pull/295
17:31:26  * mikealjoined
17:33:43  * `3rdEdenquit (Quit: Leaving...)
17:33:58  <indutny>mmalecki: one second, I'll merge v0.6 first
17:34:08  <mmalecki>indutny: can you merge v0.6 after this?
17:34:16  <mmalecki>indutny: it's technically a bugfix
17:34:31  <indutny>mmalecki: oh, k
17:34:55  <indutny>ah, sorry, I can't
17:35:02  <indutny>it's libuv, I have no access
17:35:07  <piscisaureus_>ah
17:36:18  * orlandovftwjoined
17:38:13  * paddybyersquit (Quit: paddybyers)
17:47:18  * `3rdEdenjoined
17:48:42  <isaacs>i think we should add `jslint` and `cpplint` to our `test-all` target.
17:49:18  <isaacs>it'll be painful at first, but it'd be nice to be able to actually claim that everything is consistent. i think we don't notice problems because we don't run those as often.
17:52:31  <mmalecki>is adding a self-referencing link as a test fixture acceptable?
17:52:56  <mmalecki>writing a test for ELOOP mapping, looks like this is the error we're seeing with jitsu on OS X.
17:53:09  <mmalecki>does windows even allow self-referencing links?
17:54:35  <indutny>please don't land anything in master
17:54:40  <indutny>merge in 5 minutes
17:55:58  <AndreasMadsen>isaacs: bnoordhuis tells me that he won't have jslint fixes :/
17:56:10  <isaacs>indutny: wait a second
17:56:17  <isaacs>indutny: you're merging 0.6 into master?
17:56:37  <AndreasMadsen>s/tells/told/
17:57:02  <mmalecki>AndreasMadsen: they are fine if they are reasonable
17:57:13  <isaacs>AndreasMadsen: "won't have" meaning, "they won't be done today" or meaning "I won't stand for this!"
17:57:24  <isaacs>it's ok to take time, and even to just say 0.6 isn't getting them.
17:57:40  <isaacs>but we need to make linting a part of our routine.
17:57:56  <AndreasMadsen>isaacs: meaning he won't land patches there is just code cleanup.
17:58:23  <isaacs>well, if it's jsut one or to lines, sure, whatever.
17:58:38  <isaacs>we should have one patch that lints everything, and never again accept a patch that breaks lint.
17:58:40  <isaacs>treat it as a failing test.
17:58:55  <indutny>isaacs: sorry, yes
17:58:59  <isaacs>and this should happen sometime after 0.6 is declared dead.
17:59:12  <indutny>hm...
17:59:18  <isaacs>indutny: ok, that's fine. i'm about to land doc reskin in 0.6
17:59:35  <indutny>oh, ok. I'll merge it later
17:59:41  <AndreasMadsen>isaacs: search for jslint in this log -> http://piscisaureus.no.de/log/2012-01-18
17:59:46  <isaacs>indutny: but... we need to release 0.7.1 today
17:59:51  <isaacs>indutny: so, how about you go first.
18:00:19  <isaacs>indutny: merge 0.7, i'll wait for you.
18:00:21  <AndreasMadsen>"i appreciate the effort but i don't like purely aesthetic changes"
18:01:04  <indutny>isaacs: ok
18:01:17  <isaacs>well.. that's something that has to chagne.
18:01:31  <isaacs>i don't want 100 patches from 100 different people that each fix one tiny lint nit.
18:02:09  <isaacs>indutny: just let me know if you think you have to drop anything major into master again today.
18:03:26  <isaacs>AndreasMadsen: and our jslint style is kind of insane in some cases. it's not my personal preference. but whatever. it's what we have a tool for, it's established, and it's reasonable inoffensive to most javascripters.
18:03:37  <indutny>isaacs: ok
18:03:48  <mmalecki>so, is this libuv patch going in today?
18:04:08  <mmalecki>(my first serious libuv patch, so I'm kinda overly excited :))
18:04:19  <mmalecki>well, not really serious :/
18:04:30  <AndreasMadsen>isaacs: I agree - but could the core team not find another jslint tool.
18:04:52  <mmalecki>but it actually fixes some stuff
18:05:18  <indutny>isaacs: I've connectivity problems with github, trying to push my commit (into my repo)
18:06:03  <isaacs>indutny: if you can scp a git-format-patch somewhere, that works, too
18:10:38  * AndreasMadsenchanged nick to mad_son
18:13:54  <indutny>isaacs: https://github.com/indutny/node/commit/b7c4d9d6b92f261f9c3eb0c634f4594ca8f95601
18:13:57  <indutny>review please
18:17:00  <indutny>piscisaureus_: ^
18:18:08  * CoverSlidejoined
18:19:29  <mad_son>indutny: should that be merged intro v0.7 - cluster.js looks funny
18:20:06  <indutny>mad_son: I think it should
18:20:22  * mad_sonchanged nick to AndreasMadsen
18:20:49  <AndreasMadsen>indutny: Then it is wrong
18:21:01  <indutny>AndreasMadsen: sorry?
18:22:02  <AndreasMadsen>indutny: handleWorkerMessage is not used in 0.7 and startMaster has also been removed.
18:22:26  <indutny>AndreasMadsen: oooh
18:22:33  <indutny>AndreasMadsen: ok, better do that later
18:23:19  <AndreasMadsen>indutny: actually I do not think v0.6 should change anything in cluster.js
18:23:52  <indutny>AndreasMadsen: ok, tomorrow
18:24:18  <indutny>isaacs: I won't merge today :)
18:24:47  <AndreasMadsen>I'm cool :)
18:25:18  <isaacs>AndreasMadsen, indutny: we need to get the http leak and dgram fixes into 0.7
18:25:39  <indutny>isaacs: I can ignore cluster chnages
18:25:40  <AndreasMadsen>isaacs: http leak
18:25:54  <AndreasMadsen>?
18:25:54  <isaacs>AndreasMadsen: memory leak when slamming an http proxy.
18:26:06  <CoverSlide>is that new?
18:26:13  <isaacs>AndreasMadsen: we leak parser objects pretty badly in 0.6.0 - 0.6.8
18:26:25  <isaacs>CoverSlide: it showed up in 0.5.5, iirc
18:26:30  <isaacs>CoverSlide: just found it recently, thouugh
18:26:35  <CoverSlide>oic
18:26:45  <mmalecki>oh, we do?
18:26:55  * mmaleckigoes to check memory usage at LBs
18:26:56  <isaacs>it's preventing joyent from upgrading cloud analytics to 0.6
18:27:23  <AndreasMadsen>isaacs: I know that, I thought it had been fixed. Why has cluster anything to do with http, it is just TCP port sharing.
18:27:25  <`3rdEden>^ do want that on 0.6 indeed
18:27:25  <mmalecki>our load balancers are on 0.6 already
18:28:21  <isaacs>AndreasMadsen: oh, cluster doesn't have anything to do with that.
18:28:32  <isaacs>AndreasMadsen: i'm saying, we need to have a 0.6->master merge today.
18:28:39  <isaacs>0.7.1 should get that fix.
18:28:53  <mmalecki>are you guys planning to update libuv?
18:28:56  <AndreasMadsen>isaacs: I see :)
18:29:02  <isaacs>so, we ought to not let cluster be a blocker.
18:29:14  <isaacs>we can cherry-pick around it, or exclude those patches from the merge, that's fine.
18:29:53  <isaacs>but it looks like a simple thing. probably indutny just made the wrong choice in the conflict. nothing abnormally awful :)
18:30:01  <AndreasMadsen>isaacs: The dram problem - is that about cluster not supporting UDP port sharing.
18:30:21  <isaacs>AndreasMadsen: no, that's about node throwing "Not yet implemented" from several dgram functions
18:30:34  * harryjjoined
18:30:49  <isaacs>multicast and ttl
18:31:06  <indutny>isaacs: 1 minute
18:31:13  <indutny>I just excluded lib/cluster.js changes
18:31:19  <isaacs>indutny: you are a rockstar.
18:31:23  <indutny>AndreasMadsen: is that correct ^
18:31:28  <indutny>isaacs: yes, I am :D
18:31:53  <AndreasMadsen>indutny: link
18:32:24  <AndreasMadsen>But it should be right :)
18:32:40  <isaacs>so, reviewing this patch, i'm beginning to see somewhat the wisdom of bnoordhuis's "no patches just for lint" idea.
18:32:41  <indutny>https://github.com/indutny/node/commit/667aae596cded9336f50574386683ec39ada43f2
18:32:48  <indutny>isaacs: ++
18:32:48  <kohai>isaacs has 5 beers
18:32:50  <isaacs>it does make a lot of clutter.
18:33:02  <indutny>isaacs: just do it in master, not in v0.6
18:33:10  <isaacs>so, let's stop linting 0.6, except for making sure that we don't *add* problems
18:33:34  <AndreasMadsen>jslinting hehe
18:33:37  <isaacs>then, once 0.6 is put to bed (hopefully this week, we'll know by next week for sure), then do one big push to lint the hell out of master.
18:33:50  <isaacs>otherwise it'll just be needless merge conflicts.
18:34:11  <isaacs>but, ideally, after today's fixes, and then the doc update, 0.6 won't have much more to offer us anyway very soon.
18:34:16  <indutny>isaacs: even more, it's hard to notice something wrong in that mess
18:34:21  <isaacs>indutny: yeah.
18:34:39  <CIA-115>libuv: Maciej Małecki v0.6 * r24e6c7e / (4 files in 3 dirs):
18:34:39  <CIA-115>libuv: unix: map `ENAMETOOLONG` to `UV_ENAMETOOLONG`
18:34:39  <CIA-115>libuv: With tests. Closes #295 - http://git.io/lcV6bA
18:35:05  <mmalecki>piscisaureus_: thanks!
18:35:38  <AndreasMadsen>indutny: test/simple/test-cluster-uncaught-exception.js should be excluded but we will properly survive.
18:35:52  <AndreasMadsen>There is cluster2 testcase for that
18:36:18  * travis-cijoined
18:36:18  <travis-ci>[travis-ci] joyent/libuv#48 (v0.6 - 24e6c7e : Maciej Małecki): The build is still failing.
18:36:18  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/edbabe6...24e6c7e
18:36:18  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/563958
18:36:18  * travis-cipart
18:36:20  <AndreasMadsen>Sorry has to go, but there is no other cluster issues
18:36:26  <indutny>AndreasMadsen: cool
18:36:35  <indutny>isaacs: other things lgtm, what do you think?
18:36:45  <isaacs>still reviewing.. almost don.e
18:37:40  <isaacs>yes, lgtm
18:37:43  <isaacs>ship it
18:38:00  <indutny>k
18:43:42  <indutny>isaacs: can you merge it for me?
18:43:50  <indutny>isaacs: github is refusing my connections
18:43:55  <isaacs>indutny: sure, will do.
18:43:55  <indutny>ooh
18:43:59  <CIA-115>node: Fedor Indutny master * r667aae5 / (73 files in 12 dirs): (log message trimmed)
18:43:59  <CIA-115>node: Merge branch 'v0.6'
18:43:59  <CIA-115>node: Conflicts:
18:43:59  <CIA-115>node: ChangeLog
18:43:59  <CIA-115>node: doc/template.html
18:44:00  <CIA-115>node: lib/cluster.js
18:44:00  <CIA-115>node: lib/http.js
18:44:00  <indutny>looks like it accepted one
18:44:02  <indutny>:D
18:44:04  <indutny>heheh
18:44:05  <isaacs>oh, heh
18:44:06  <isaacs>nvm
18:44:09  <isaacs>thanks, indutny
18:48:36  <bnoordhuis>indutny: what was the conflict in http.js?
18:48:58  <bnoordhuis>the diff looks okay btw, just curious
18:50:09  <mmalecki>prolly deprecated thingie
18:50:35  <bnoordhuis>it's a little too big for that
18:50:42  * bnoordhuisintroduces a twss moment
18:50:48  <mmalecki>it's the only change I can remember not going into v0.6 and going into master...
18:50:49  <mmalecki>lol
18:51:33  <isaacs>bnoordhuis: hey
18:51:42  <isaacs>bnoordhuis: care to scroll up and review discussion of linting?
18:52:00  <isaacs>bnoordhuis: your opinion was tossed around without you being here to confirm or deny anything :)
18:52:47  <bnoordhuis>isaacs: right, AndreasMadsen more or less has it right - i'm not a fan of big linting patches
18:53:14  <bnoordhuis>merge conflicts, doesn't necessarily make the code more readable, decreases the effectiveness of `git blame`, etc.
18:53:14  <isaacs>bnoordhuis: yes, they can be a mess.
18:53:20  <isaacs>yeah
18:53:26  <isaacs>but, they have benefits as well.
18:54:03  <isaacs>i think once we put 0.6 to bed, we should just bite the bullet and make it a policy moving forward. one big commit to lint everything, then never again accept a linty patch
18:54:05  <mmalecki>coffeescript is nicer, we should use it.
18:54:27  <isaacs>./kick mmalecki
18:54:32  <bnoordhuis>i also hear good things about java
18:54:32  <isaacs>;P
18:54:41  <mmalecki>coco is also good.
18:54:41  <isaacs>: you guys are jerks.
18:54:51  <bnoordhuis>isaacs: the most important thing to me is readability
18:54:58  <isaacs>bnoordhuis: yes. that is the most important thing.
18:55:03  <bnoordhuis>and the node source is - by and large - pretty readable
18:55:07  <isaacs>the second most important thing is consistency.
18:55:18  <isaacs>linting isn't for you or me. you know my feelings about style.
18:55:38  <isaacs>it's more about a perception of making node a solid and stable platform.
18:55:40  <mmalecki>I don't understand how can linting increase readability
18:55:47  <bnoordhuis>ah well, i won't fight it (very hard)
18:55:59  <bnoordhuis>it's just that i don't see the need
18:56:01  <isaacs>bnoordhuis: i fought hard when ryan introduced the linter in the first place.
18:56:03  * orlandovftwquit (Ping timeout: 255 seconds)
18:56:22  <isaacs>bnoordhuis: but as people whine about "use strict" and other bullshit, we need to have an answer that isn't "it's fine, trust us, really"
18:56:27  <isaacs>even though, of course, that's the only sane answer.
18:56:35  <isaacs>it's not convincing to people who eat the fud.
18:57:14  <bnoordhuis>let them
18:57:16  <isaacs>plus, moving forward, consistency does tend to increase readability
18:57:18  <bnoordhuis>you can't please everyone
18:57:41  <isaacs>bnoordhuis: no, but we can please a lot of decision makers in a lot of places and also make our codebase a little more consistent.
18:57:54  <isaacs>i just don't want to do it in 0.6, or in a way that's going to just be a mess of conflicts the next time we merge.
18:57:54  * travis-cijoined
18:57:54  <travis-ci>[travis-ci] joyent/node#304 (master - 667aae5 : Fedor Indutny): The build is still failing.
18:57:54  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/2433eeb...667aae5
18:57:54  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/564000
18:57:54  * travis-cipart
18:58:03  <isaacs>or have 100 "lint" commits all over the place.
18:58:18  <bnoordhuis>i would hate that even more so okay
18:58:32  <isaacs>ok.
18:59:04  <isaacs>so, my plan is, after 0.7.2, but before 0.7.4, one commit to lint them all, and we add jslint and cpplint to our test-all make target
18:59:12  <isaacs>or maybe just to the test target.
18:59:39  <isaacs>and never ever discuss this again.
18:59:41  <isaacs>:)
18:59:45  <AndreasMadsen>:)
18:59:57  <AndreasMadsen>can we clear the log :)
18:59:59  <isaacs>part of the point is that style is so distracting, too.
19:00:33  <isaacs>it's like naming "node_modules". like, no, you can't call it "vendor" or "deps" or whatever else. we're picking a name that's blessed. no one likes it. eat your vegetables and shut up.
19:00:55  <mmalecki>but... but...!
19:01:30  <mmalecki>I want to call it 'gems'.
19:02:51  <bnoordhuis>isaacs, piscisaureus_: what are the remaining big issues for 0.6?
19:02:57  <AndreasMadsen>Oh it is do difficult to find a good module name these days.
19:03:11  <isaacs>bnoordhuis: i'm going to land a pretty significant doc patch soon.
19:03:19  <isaacs>bnoordhuis: but i think that's about it
19:03:45  <isaacs>not doc content, just style
19:04:10  <isaacs>heading into sf, i'll be back in a little bit
19:04:25  * isaacsquit (Remote host closed the connection)
19:04:52  <piscisaureus_>bnoordhuis: eh, file watchers. udp multicast stuff.
19:05:13  * dshaw_joined
19:05:28  <bnoordhuis>piscisaureus_: didn't igor implement multicast?
19:05:39  <piscisaureus_>oh did that land already?
19:05:45  <bnoordhuis>https://github.com/joyent/libuv/commit/726e36c
19:05:59  <piscisaureus_>oh who reviewed that?
19:06:06  <bnoordhuis>uhm... igor?
19:06:29  <bnoordhuis>probably still needs support for joining/leaving multicast groups
19:06:52  <piscisaureus_>bnoordhuis: we should support the full gamut of 0.4 api
19:07:05  <bnoordhuis>piscisaureus_: 'we' is 'us windows guys'?
19:07:17  <piscisaureus_>bnoordhuis: also unix
19:07:21  <bnoordhuis>i think we do now
19:07:30  <piscisaureus_>bnoordhuis: I think there is some multicast loopback option that we don't support yet?
19:07:44  <CIA-115>libuv: Brandon Benvie v0.6 * r3ade5f0 / src/win/tty.c : Windows: Add support for interpreting axiterm-style colors. - http://git.io/XgqySQ
19:08:05  <bnoordhuis>piscisaureus_: oh right... do people actually use that?
19:08:13  <bnoordhuis>no one's complained about that yet afaik
19:08:15  <piscisaureus_>bnoordhuis: *shrug* you wanna remove it?
19:08:27  <AndreasMadsen>indutny: could you put the disconnect patch on you todo list for tomorrow or some other day.
19:08:42  <bnoordhuis>piscisaureus_: i want to say yes...
19:08:52  <piscisaureus_>bnoordhuis: http://nodejs.org/docs/v0.4.12/api/dgram.html#dgram.setMulticastLoopback
19:09:13  <AndreasMadsen>https://github.com/joyent/node/pull/2591 - it blocks some cluster work :)
19:09:17  <bnoordhuis>piscisaureus_: yeah, i know what you mean
19:09:24  <bnoordhuis>oh, we're still missing plain ttl support
19:09:26  * travis-cijoined
19:09:26  <travis-ci>[travis-ci] joyent/libuv#49 (v0.6 - 3ade5f0 : Brandon Benvie): The build is still failing.
19:09:26  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/24e6c7e...3ade5f0
19:09:26  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/564151
19:09:26  * travis-cipart
19:09:29  <piscisaureus_>bnoordhuis: it is actually also documented for later versions
19:09:38  <piscisaureus_>just not supported
19:09:49  <piscisaureus_>so it has to be fixed either way - or properly remove it, or add it again
19:10:04  <mmalecki>piscisaureus_: hey, can you do circular symlink on windows?
19:10:05  <bnoordhuis>let's just add it, it's trivial
19:10:19  <piscisaureus_>mmalecki: sure you can
19:10:54  <mmalecki>piscisaureus_: ok. I'm writing test for ELOOP mapping and just wondering if that's actually possible.
19:11:12  <mmalecki>does it sound sane to put a circular link as a text fixture there?
19:11:18  <mmalecki>or is it better to create at runtime?
19:11:35  <bnoordhuis>mmalecki: we actually cut down the number of errnos a while ago...
19:12:10  <mmalecki>bnoordhuis: why?
19:12:30  <piscisaureus_>mmalecki: atm it results in UNKNOW error
19:12:39  <mmalecki>piscisaureus_: I know
19:12:45  <mmalecki>we've seen it with jitsu.
19:13:21  <mmalecki>it's not like I'm grabbing random errnos and mapping them.
19:14:09  <piscisaureus_>mmalecki: while you're at it you might wanna add a mapping for WSAENOBUFS
19:14:14  <piscisaureus_>(for windows)
19:15:24  <mmalecki>piscisaureus_: it's like EMFILE?
19:15:39  <piscisaureus_>mmalecki: no just map it to UV_ENOBUFS
19:15:45  <piscisaureus_>I think we have it
19:15:52  <mmalecki>piscisaureus_: I don't think I can do that, tho, I have no Windows machine and I'm not planning to have one ;)
19:16:13  <bnoordhuis>hey, the rap negro is walking past my house again
19:16:23  <mmalecki>or well, maybe. Battlefield is pretty good actually
19:16:25  <bnoordhuis>big black man who's always rapping out aloud, hence 'rap negro'
19:17:29  <mmalecki>lol
19:17:52  <mmalecki>damn, now I want to play battlefield :/
19:21:01  <bnoordhuis>mmalecki: real men play quake 3, fact
19:21:09  <mmalecki>bnoordhuis: openarena ftw
19:21:14  * brsonjoined
19:21:17  <bnoordhuis>very good
19:21:22  <bnoordhuis>(contains patches by yours truly)
19:21:49  <mmalecki>bnoordhuis: really? that's neat!
19:22:23  <CIA-115>libuv: Bert Belder v0.6 * r6554954 / src/win/error.c : Windows: add error mapping for WSAENOBUFS - http://git.io/uGHNmw
19:22:36  <mmalecki>piscisaureus_: ++
19:22:37  <kohai>piscisaureus_ has 4 beers
19:22:59  <ryah_>we need to fix EMFILE for tcp servers at some point
19:23:06  <ryah_>it should back off accepting new connections
19:23:19  <ryah_>and do that trick to reject new ones
19:24:06  * travis-cijoined
19:24:06  <travis-ci>[travis-ci] joyent/libuv#50 (v0.6 - 6554954 : Bert Belder): The build is still failing.
19:24:06  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/3ade5f0...6554954
19:24:06  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/564241
19:24:06  * travis-cipart
19:27:47  <piscisaureus_>ryah_: that trick as very nice but when you reach the FD limit you are already fucked
19:27:53  <piscisaureus_>because you cannot open a file etc
19:28:06  <piscisaureus_>ryah_: it would be better to be able to pause the server
19:29:39  <bnoordhuis>hah, i was thinking about adding the EMFILE hack again
19:29:47  <bnoordhuis>today, that is
19:29:51  <piscisaureus_>unix guys :-/
19:30:12  <piscisaureus_>fd limits are so retarded
19:30:54  <bnoordhuis>but they're here and we have to live with them
19:31:21  <piscisaureus_>Unix, we are getting a divorce
19:31:35  <ryah_>"Node to not support Unix systems anymore"
19:31:53  <piscisaureus_>Not that windows is any better
19:31:58  <piscisaureus_>we should support BeOS
19:32:11  <mmalecki>haiku os :3
19:32:18  <ryah_>plan9
19:32:28  <bnoordhuis>hah, i was just about to say that
19:33:00  <bnoordhuis>plan9, for that ultra smug feeling
19:33:10  <mmalecki>but really, haiku is lovely. I kinda want it to become popular.
19:33:42  <ryah_>do you guys want to see the most rad thing ever
19:33:43  <ryah_>http://info.iet.unipi.it/~luigi/netmap/
19:33:57  <bnoordhuis>mmalecki: that ship has sailed - sometime in 1999 actually
19:34:11  <mmalecki>bnoordhuis: :(
19:34:23  <mmalecki>ok, hey, I've got an OS we should totally support!
19:34:35  <mmalecki>https://github.com/mmalecki/cOS
19:34:36  <ryah_>exposes nic ring buffer safely via mmapped region
19:36:07  <bnoordhuis>ryah_: looks hip
19:36:13  <piscisaureus_>mmalecki: Haiku is too c++ y for me
19:36:16  <bnoordhuis>at what level of the stack does it work?
19:36:28  <ryah_>bnoordhuis: it's just raw packets
19:36:30  <bnoordhuis>raw packets?
19:36:31  <bnoordhuis>right
19:36:33  <piscisaureus_>mmalecki: Imho a good kernel exposes only a raw syscall interface.
19:36:38  <ryah_>so you'd need to write a tcp stack
19:36:54  <ryah_>but it gets a packet from wire into the app in 90 clock cycles
19:36:59  <ryah_>(supposedly)
19:37:06  <ryah_>which makes my penis hard
19:37:36  <mmalecki>piscisaureus_: hey, I was 15 when writing it!
19:37:44  <mmalecki>it was supposed to be the best OS ever.
19:38:57  <bnoordhuis>ryah_: it's in freebsd 9? have you played with it yet?
19:38:57  <harryj>http://www.youtube.com/watch?v=SPtoXNW9yEQ
19:41:31  <ryah_>bnoordhuis: not yet
19:41:43  <piscisaureus_>mmalecki: the windows code for ELOOP is ERROR_CANT_RESOLVE_FILENAME
19:42:01  <mmalecki>piscisaureus_: oh, cool, thanks!
19:42:08  * isaacsjoined
19:42:12  <mmalecki>(I guess it isn't exact equivalent, right?)
19:42:57  <piscisaureus_>mmalecki: sure but who cares
19:43:07  <mmalecki>piscisaureus_: ++
19:43:07  <kohai>piscisaureus_ has 5 beers
19:43:35  <mmalecki>piscisaureus_: I'm done with mapping, just writing a test case (also, standup). I'll be done today.
19:46:47  <piscisaureus_>mmalecki: afaict ERROR_CANT_RESOLVE_FILENAME is generated when a symlink is not broken but the kernel reports STATUS_REPARSE_POINT_NOT_RESOLVED anyway. I can't imagine any other situation where that might happen other than ELOOP.
19:47:55  <mmalecki>piscisaureus_: reparse points are a bit more generic than symlinks in unix, right?
19:48:10  <piscisaureus_>mmalecki: yeah.
19:48:22  <mmalecki>but again, meh :)
19:52:46  <harryj>Does anyone contributing to Node need to sign: http://nodejs.org/cla.html ?
19:52:56  <isaacs>harryj: yes.
19:52:56  <mmalecki>harryj: yes
19:54:55  * orlandovftwjoined
20:02:27  <piscisaureus_>ryah_: the netmap thing is cool but I dislike the fact that it requires people to run as root. You might get away with in for RyOS but for BertOS I am looking for a little more safety.
20:03:34  <piscisaureus_>I like the idea though
20:04:28  <piscisaureus_>I wonder if it would be possible for the kernel to remove a part of the ring buffer and add it to the user's address space
20:04:38  <piscisaureus_>that would solve it
20:07:28  * orlandovftwquit (Quit: leaving)
20:08:01  * orlandovftwjoined
20:09:21  <harryj>You already need to run as root for ports < 1024. I don't see the issue, ie you can create an appplication tha provides access to the area safely.
20:25:27  * perezdjoined
20:26:03  * mikealquit (Quit: Leaving.)
20:27:46  * mikealjoined
20:33:40  * mralephjoined
20:41:18  * AndreasMadsenquit (Remote host closed the connection)
20:50:11  * piscisaureus_quit (Ping timeout: 240 seconds)
21:09:46  * piscisaureus_joined
21:12:26  <CIA-115>node: isaacs v0.6 * r5c0d11b / Makefile : doc: Add images to makefile - http://git.io/KiZ1HA
21:12:28  <CIA-115>node: isaacs v0.6 * r6768d2f / (9 files in 6 dirs): doc: Pixel-nudging - http://git.io/uh7Wpw
21:12:29  <CIA-115>node: isaacs v0.6 * r5d7577c / tools/doctool/doctool.js : doctool: correct improperly nested ul/li handling - http://git.io/4R4gsA
21:12:30  <CIA-115>node: isaacs v0.6 * r4fd3151 / (13 files in 2 dirs): Add images to doc_assets - http://git.io/eXVgOA
21:12:31  <CIA-115>node: Matthew Fitzsimmons v0.6 * r831c819 / (9 files in 5 dirs): Nodejs.org design refresh. - http://git.io/42KGKA
21:14:32  * piscisaureus__joined
21:14:39  * piscisaureus_quit (Ping timeout: 252 seconds)
21:18:04  * `3rdEdenquit (Ping timeout: 272 seconds)
21:20:08  * travis-cijoined
21:20:08  <travis-ci>[travis-ci] joyent/node#305 (v0.6 - 6768d2f : isaacs): The build passed.
21:20:08  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/6c0c00a...6768d2f
21:20:08  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/564717
21:20:08  * travis-cipart
21:20:28  <isaacs>docs *always* fix the build.
21:20:31  <isaacs>:)
21:20:50  <mmalecki>fork2 is race-y, we shouldn't kill the worker there
21:42:52  * `3rdEdenjoined
21:45:24  * piscisaureus_joined
21:45:52  * piscisaureus_quit (Client Quit)
21:51:02  <CIA-115>node: isaacs v0.6 * r7b93eb0 / doc/api_assets/style.css : doc: Remove default border around logo for MSIE - http://git.io/ECbjAg
21:57:57  <isaacs>ok, last chance before pushing live, anyone wanna try to find some bugs?
21:57:58  <isaacs>http://static.izs.me/nodejs.org-reskin-2012/api/
21:58:03  <isaacs>http://static.izs.me/nodejs.org-reskin-2012/about/
21:58:05  <isaacs>http://static.izs.me/nodejs.org-reskin-2012/
21:58:07  <isaacs>http://static.izs.me/nodejs.org-reskin-2012/community/
21:58:26  <isaacs>note that the links don't work, since it's not in the root on static.izs.me
21:58:30  <isaacs>but it is on nodejs.org
21:58:50  <CIA-115>node: isaacs v0.6 * rf33a35e / (5 files in 4 dirs): doc: trademark link must be absolute - http://git.io/PVEDbw
21:59:42  * travis-cijoined
21:59:42  <travis-ci>[travis-ci] joyent/node#306 (v0.6 - 7b93eb0 : isaacs): The build passed.
21:59:42  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/6768d2f...7b93eb0
21:59:42  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/564847
21:59:42  * travis-cipart
22:03:51  <bnoordhuis>http://static.izs.me/nodejs.org-reskin-2012/ <- does it need this newfangled javascript?
22:04:44  <isaacs>bnoordhuis: newfangled?
22:04:53  <isaacs>bnoordhuis: you mean the #download click handler stuff?
22:05:14  <bnoordhuis>it was a (half) tongue in cheek jest
22:05:45  <isaacs>bnoordhuis: i agree, it's kind of terribel.
22:05:49  <bnoordhuis>i take it the big download popup is gone?
22:05:54  <isaacs>bnoordhuis: no, it's still there
22:05:56  <isaacs>click "download"
22:06:00  <bnoordhuis>oh wait, indeed
22:06:12  <bnoordhuis>my newfangled javascript browser took a while to catch on
22:06:12  <isaacs>bnoordhuis: yeah, i think changing that will take more effort.
22:06:36  * travis-cijoined
22:06:36  <travis-ci>[travis-ci] joyent/node#307 (v0.6 - f33a35e : isaacs): The build passed.
22:06:36  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/7b93eb0...f33a35e
22:06:36  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/564886
22:06:36  * travis-cipart
22:07:19  <bnoordhuis>community page looks good
22:09:35  <isaacs>what about the docs page?
22:09:49  <isaacs>bnoordhuis: http://static.izs.me/nodejs.org-reskin-2012/api/child_processes.html
22:10:40  <bnoordhuis>isaacs: a lot purdier than what we have now
22:10:53  <isaacs>kewl
22:10:59  <isaacs>k, i'm gonna push it live.
22:17:47  <`3rdEden>isaacs awesomeness
22:17:58  <mmalecki>isaacs: ++
22:17:58  <kohai>isaacs has 6 beers
22:18:12  <`3rdEden>isaacs except for the up arrow for home and the down arrow for download
22:18:16  <`3rdEden>that's a bit confusing..
22:18:26  <CoverSlide>the white hurts my eyes!
22:19:29  <`3rdEden>OMG version number on the front page ;D
22:19:38  <`3rdEden>isaacs:++
22:19:39  <kohai>isaacs has 7 beers
22:19:40  <`3rdEden>;D
22:21:03  * `3rdEdenquit (Quit: Zzzz of the mission bay for prep work)
22:21:27  <isaacs>let's not start bikeshedding the design.
22:21:32  <isaacs>or at least, don't complain to me about the design :)
22:22:00  <isaacs>i helped make the pixels on the webpage match the pixels on the image, but even that was mostly matt fitzsimmons.
22:37:05  * mralephquit (Quit: Leaving.)
23:00:11  <isaacs>anyone see anything missing? https://gist.github.com/ce6bc4f8e865c663c7e0
23:01:07  <mmalecki>isaacs: did my uv changes go in?
23:01:15  <mmalecki>I guess not
23:01:17  <isaacs>mmalecki: which ones?
23:01:24  <isaacs>mmalecki: i think they were on 0.6 already, yes?
23:01:38  <mmalecki>indutny: yes, on v0.6 branch of libuv
23:01:56  <mmalecki>* isaacs ^
23:02:05  <isaacs>mmalecki: i see
23:02:13  <mmalecki>they didn't. meh, it's not that important
23:02:17  <isaacs>yeah, i dind't mention the changes that will already covered in 0.6 changelogs.
23:09:15  <CIA-115>libuv: Ben Noordhuis v0.6 * r02b41f6 / include/uv.h : Add missing UV_EXTERN declarations in uv.h - http://git.io/SdvWTw
23:09:15  <CIA-115>libuv: Ben Noordhuis v0.6 * re710fdb / (include/uv.h src/unix/udp.c src/win/udp.c): unix: implement uv_udp_set_ttl() - http://git.io/Tr-EXA
23:09:16  <CIA-115>libuv: Ben Noordhuis v0.6 * rb88bc43 / (include/uv.h src/unix/udp.c src/win/udp.c): unix: implement uv_udp_set_multicast_loop() - http://git.io/tE0Ktg
23:09:47  <piscisaureus__>bnoordhuis: ^-- who reviewed that?
23:10:05  <bnoordhuis>piscisaureus__: uhm... me?
23:10:18  <bnoordhuis>the win implementation simply raises ENOSYS
23:10:56  * travis-cijoined
23:10:57  <travis-ci>[travis-ci] joyent/libuv#51 (v0.6 - b88bc43 : Ben Noordhuis): The build is still failing.
23:10:57  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/6554954...b88bc43
23:10:57  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/565257
23:10:57  * travis-cipart
23:12:25  <CIA-115>node: Ben Noordhuis v0.6 * r2775c0e / (3 files in 3 dirs): dgram: bring back setTTL() (+8 more commits...) - http://git.io/5D9-jw
23:13:04  <bnoordhuis>^ lots of bug fixes in that one
23:13:39  * sh1mmer_joined
23:13:41  * sh1mmer_quit (Remote host closed the connection)
23:14:23  * sh1mmer_joined
23:15:00  <bnoordhuis>piscisaureus__: pig
23:15:04  <bnoordhuis>heh, ping
23:15:07  <bnoordhuis>sorry, typo
23:16:07  * sh1mmer_quit (Client Quit)
23:17:05  * sh1mmerquit (Ping timeout: 245 seconds)
23:20:10  * travis-cijoined
23:20:10  <travis-ci>[travis-ci] joyent/node#308 (v0.6 - 2775c0e : Ben Noordhuis): The build passed.
23:20:10  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/f33a35e...2775c0e
23:20:10  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/565282
23:20:10  * travis-cipart
23:20:56  <piscisaureus__>bnoordhuis: en bedankt
23:21:11  <bnoordhuis>piscisaureus__: graag gedaan
23:21:26  <bnoordhuis>so there's an issue with the broadcast/multicast/ttl functions
23:21:40  <bnoordhuis>they don't work unless you call bind first
23:21:48  <piscisaureus__>bnoordhuis: aha
23:21:59  <piscisaureus__>bnoordhuis: so what can we do about it?
23:22:09  <bnoordhuis>piscisaureus__: multiple things
23:22:38  <bnoordhuis>piscisaureus__: https://github.com/bnoordhuis/libuv/commit/81c12dd <- one possible solution
23:22:44  <bnoordhuis>gnarly though it is
23:22:45  <isaacs>piscisaureus__: so you guys decided to go live today because it's "node announcement day"?
23:22:48  <isaacs>;P
23:24:02  <bnoordhuis>piscisaureus__: another possible solution is to do an implicit bind()
23:24:17  <bnoordhuis>piscisaureus__: the best solution would be to just create the bloody socket in uv_udp_init() of course
23:24:29  <piscisaureus__>bnoordhuis: how did node 0.4 do it?
23:24:42  <piscisaureus__>bnoordhuis: ok let's get rid of the deferred socket creation
23:25:11  <piscisaureus__>bnoordhuis: but I forgot to mention - please get your stuff reviewed.
23:25:32  <piscisaureus__>(this should be isaacs job :-))
23:25:43  <bnoordhuis>piscisaureus__: oh, sure
23:26:05  <bnoordhuis>most of those commits were cleanup after a PR i merged this weekend though
23:26:14  <piscisaureus__>isaacs: I don't know why we went online today
23:26:26  <isaacs>bnoordhuis: yes, even still, please have someone review the code who had no part in writing it.
23:26:27  <piscisaureus__>bnoordhuis: aha - alright. then it's fine
23:26:34  <isaacs>piscisaureus__--
23:26:35  <kohai>piscisaureus__ has -1 beer
23:26:39  <isaacs>don't make excuses!
23:26:54  <isaacs>it's not a big deal in this case.
23:27:07  <piscisaureus__>isaacs: Actually I think we were planning to go online a week ago but there were some issues
23:27:11  <piscisaureus__>isaacs: wut?
23:27:19  <piscisaureus__>isaacs: why are you -- ing me ???
23:27:31  <isaacs>piscisaureus__: because you said it's fine to not have stuff reviewe.
23:27:35  <isaacs>*reviewed
23:27:40  <isaacs>here, you can have your beer back:
23:27:42  <isaacs>piscisaureus__++
23:27:42  <kohai>piscisaureus__ has 0 beer
23:27:46  <isaacs>piscisaureus__++
23:27:48  <kohai>piscisaureus__ has 1 beer
23:27:52  <piscisaureus__>isaacs: you are only blaming the messenger :-/
23:27:59  <isaacs>fair enough.
23:28:49  <piscisaureus__>now I don't want it anymore (weeks)
23:28:49  <piscisaureus__>piscisaureus--
23:28:50  <kohai>piscisaureus has 0 beer
23:28:59  <bnoordhuis>bnoordhuis++
23:28:59  <piscisaureus__>*weeps
23:29:00  <kohai>You can't give karma to yourself!
23:29:04  <bnoordhuis>hrm
23:29:15  <mmalecki>:)
23:29:22  <mmalecki>I implemented beer plugin, btw.
23:29:24  * mmaleckiducks
23:29:39  * piscisaureus__changed nick to asdjasdasdad
23:29:45  <asdjasdasdad>piscisaureus++
23:29:46  <mmalecki>yeah, that works.
23:29:46  <kohai>piscisaureus has 1 beer
23:29:51  * asdjasdasdadchanged nick to piscisaureus
23:29:52  <bnoordhuis>piscisaureus__: so anyway.... yes, let's create the socket immediately
23:30:02  <bnoordhuis>piscisaureus: but that won't fly for 0.6
23:30:02  <piscisaureus>bnoordhuis++
23:30:04  <kohai>bnoordhuis has 5 beers
23:30:12  <bnoordhuis>that many huh? good
23:30:17  <piscisaureus>hmpf
23:30:30  <piscisaureus>bnoordhuis: why does that not work for 0.6? because libuv can't change api?
23:30:34  <bnoordhuis>yes
23:30:45  <piscisaureus>fair enough
23:30:54  <piscisaureus>bnoordhuis: maybe we could just cache the flags is JS land
23:31:26  <bnoordhuis>the downside to that approach is that you don't get errors until (possibly much) later
23:31:35  <piscisaureus>I know
23:31:43  <piscisaureus>how hard is it to implement
23:31:55  <bnoordhuis>not hard, it's just not very reliable / clean
23:31:56  <piscisaureus>besides, how likely are these options to fail anyway?
23:32:05  <piscisaureus>yes I know
23:32:11  <piscisaureus>let's make this better in 0.7
23:32:12  <bnoordhuis>maybe we can update the docs instead: bind() before you call this function
23:32:26  <bnoordhuis>and fix it in 0.7, yes
23:32:38  <piscisaureus>also fine
23:33:11  <isaacs>test please: http://nodejs.org/dist/v0.7.1/node-v0.7.1-RC0.tar.gz
23:33:20  <isaacs>(building windows now)
23:37:23  <isaacs>windows build failing: https://gist.github.com/1666423
23:37:26  <isaacs>piscisaureus: any ideas? ^
23:38:58  <piscisaureus>Did you forget a ';'
23:39:11  <piscisaureus>bnoordhuis: that's what you get from not requesting review ^ :-p
23:39:18  * piscisaureussnarky comment mode off
23:39:41  <bnoordhuis>piscisaureus: that's from the PR i mean
23:40:05  <bnoordhuis>it compiles for me though
23:40:24  <isaacs>bnoordhuis: on windows?
23:40:31  <isaacs>bnoordhuis: it compiles fine for me on linux and os x
23:40:42  <bnoordhuis>sorry, that's what i mean
23:40:46  <piscisaureus>oh
23:40:54  <piscisaureus>isaacs: interface may be a reserved keyword
23:41:15  <piscisaureus>isaacs: try replacing all the instances of interface by something else
23:41:26  <bnoordhuis>piscisaureus: wut? is that a ms thing?
23:41:37  <piscisaureus>(sorry my internet is not fast enough to fetch and try. Unless you guys can wait for half an hour)
23:41:52  <isaacs>piscisaureus: lemme try that, hold on
23:43:04  <isaacs>s/<interface/iface/g seems to have fixed it
23:43:36  <isaacs>still a jillion yellow warnings, but that's normal
23:43:53  <piscisaureus>bnoordhuis: http://msdn.microsoft.com/en-us/library/737cydt1.aspx
23:44:10  <piscisaureus>isaacs: yes we should deal with them some day. I can confirm they are mostly harmless
23:44:32  <isaacs>alright, that compiles. lemme make a commit for that and have you review it.
23:44:34  <bnoordhuis>piscisaureus: but why is the c++ compiler about that?
23:44:40  <bnoordhuis>*complaining
23:45:28  <isaacs>piscisaureus: https://gist.github.com/1666533
23:45:30  <piscisaureus>bnoordhuis: yeah, good question. I will email steve.
23:45:47  <isaacs>piscisaureus: you mean you don't have his phone number?
23:45:50  <piscisaureus>isaacs: lgtm
23:45:52  <bnoordhuis>piscisaureus: tell him i said hi and "developers, developers, developers"
23:45:59  <isaacs>lol
23:46:03  <isaacs>k, moving forward...
23:46:49  <CIA-115>node: isaacs master * rb1b16d1 / src/udp_wrap.cc : udp: 'interface' is reserved word on windows. - http://git.io/Z4GpQA
23:50:01  <bnoordhuis>sorry guys, i have to go - need to get up early tomorrow
23:50:06  <bnoordhuis>good luck with the release
23:50:30  * bnoordhuisquit (Quit: Leaving)
23:53:07  * benviequit