00:00:13  * piscisaureus_quit (Read error: Connection reset by peer)
00:02:36  * bradleymeckjoined
00:05:48  * ericktquit (Quit: erickt)
00:26:15  * dylangjoined
00:32:44  * toothrjoined
00:35:50  <isaacs>grawlx.
00:35:55  <isaacs>have to update npm *again*
00:37:44  * ericktjoined
00:39:05  * bradleymeckquit (Quit: bradleymeck)
00:39:35  <CIA-131>node: isaacs v0.8 * r2f1985c / (101 files in 7 dirs): npm: Upgrade to 1.1.57 - http://git.io/ODzHdQ
00:42:04  <TooTallNate>isaacs: one-char change, huh ^? haha
00:44:38  <isaacs>TooTallNate: an important char!
00:44:45  <isaacs>2x Infinity important!
00:44:56  <TooTallNate>haha, yup
00:48:28  <isaacs>pop quiz: -Infinity + Infinity
00:53:56  <TooTallNate>i had a feeling it'd be NaN
00:54:12  * brsonquit (Ping timeout: 256 seconds)
00:54:12  <TooTallNate>go go JS!
00:54:29  * mikealjoined
00:54:48  * dapquit (Quit: Leaving.)
00:54:51  * ericktquit (Quit: erickt)
00:57:38  <isaacs>w00t!
00:58:00  * isaacsgives TooTallNate a gold star
00:58:04  <isaacs>and with that,
00:58:06  * isaacsaway
00:59:05  * ericktjoined
01:05:40  * ericktquit (Ping timeout: 252 seconds)
01:06:36  * mikealquit (Quit: Leaving.)
01:11:06  * brsonjoined
01:11:25  * joshthecoderquit (Quit: Leaving...)
01:27:23  * c4miloquit (Remote host closed the connection)
01:36:57  * abraxasjoined
01:49:54  * lohkeyquit (Quit: lohkey)
01:58:47  * ericktjoined
02:06:29  * dylangquit (Quit: dylang)
02:10:47  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
02:17:07  * joshthecoderjoined
02:22:51  * dylangjoined
02:26:59  * blackorzarquit (Remote host closed the connection)
02:39:37  * dshaw_quit (Quit: Leaving.)
02:47:04  * brsonquit (Ping timeout: 240 seconds)
03:05:55  * perezdjoined
03:07:48  * perezdquit (Client Quit)
03:19:53  * txdvjoined
03:20:58  * txdv_quit (Ping timeout: 252 seconds)
03:23:57  * c4milojoined
03:26:08  * brsonjoined
03:26:27  * mikealjoined
03:32:11  * txdvquit (Ping timeout: 265 seconds)
03:33:04  * txdvjoined
03:33:26  * ryahquit (Ping timeout: 252 seconds)
03:34:14  * dylangquit (Quit: dylang)
03:36:24  * ryahjoined
03:37:19  * AvianFlujoined
04:01:40  * ryahquit (Ping timeout: 252 seconds)
04:02:19  * ryahjoined
04:03:50  * AvianFluquit (Ping timeout: 249 seconds)
04:06:58  * mikealquit (Quit: Leaving.)
04:32:42  * bradleymeckjoined
04:55:32  * c4miloquit (Remote host closed the connection)
04:57:24  * c4milojoined
05:02:31  <indutny>hai
05:08:16  * c4miloquit (Remote host closed the connection)
05:11:49  * gicodequit (Ping timeout: 252 seconds)
05:12:35  * gicodejoined
05:18:53  * TheJHjoined
06:02:49  * ArmyOfBrucequit (Excess Flood)
06:03:20  * ArmyOfBrucejoined
06:03:40  * dshaw_joined
06:11:39  * dshaw_quit (Quit: Leaving.)
06:29:12  <ryah>hi
06:29:50  * AvianFlujoined
06:35:15  * AvianFluquit (Ping timeout: 245 seconds)
06:43:06  * brsonquit (Ping timeout: 265 seconds)
06:59:26  * loladirojoined
06:59:36  * stephankquit (Quit: *Poof!*)
07:02:42  * rendarjoined
07:03:48  * loladiroquit (Read error: Connection reset by peer)
07:04:23  * loladirojoined
07:31:24  * bnoordhuisjoined
07:32:14  * `3rdEdenjoined
07:36:13  <CIA-131>node: Ron Korving v0.8 * r2e1f2b5 / doc/api/cluster.markdown : doc: fix typo in cluster example code - http://git.io/_WW-5w
07:40:25  <bnoordhuis>indutny: re #3661, can it somehow be tested?
07:40:31  <bnoordhuis>in an automated fashion, i mean
07:49:07  <CIA-131>libuv: Ben Noordhuis master * r29eb460 / src/unix/linux/linux-core.c : linux: fix fscanf() -Wunused-result warning - http://git.io/e4NMzw
07:50:53  * travis-cijoined
07:50:53  <travis-ci>[travis-ci] joyent/libuv#607 (master - 29eb460 : Ben Noordhuis): The build passed.
07:50:53  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/03bdfd08ccc7...29eb460d0b20
07:50:53  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2197915
07:50:53  * travis-cipart
08:25:05  <CIA-131>node: isaacs v0.8.8-release * r85161a8 / (101 files in 6 dirs): npm: upgrade to 1.1.58 - http://git.io/iObRZA
08:25:06  <CIA-131>node: isaacs v0.8.8-release * rb8fcde4 / (AUTHORS ChangeLog src/node_version.h): wip - http://git.io/8-2Shg
08:29:35  * mikealjoined
08:32:13  * loladiroquit (Ping timeout: 245 seconds)
08:35:36  * AvianFlujoined
08:40:45  <CIA-131>node: Ben Noordhuis master * r1867511 / common.gypi : build: compile with -Wextra -Wno-unused-parameter - http://git.io/BIVIAw
08:46:20  * hzjoined
08:58:28  * joshthecoderquit (Quit: Linkinus - http://linkinus.com)
09:32:22  * AvianFluquit (Quit: AvianFlu)
09:45:17  * mmaleckijoined
09:47:05  <indutny>bnoordhuis: hi
09:47:10  <bnoordhuis>indutny: ho
09:47:23  <indutny>bnoordhuis: em... I think we already have tls session resumption test
09:47:34  <indutny>bnoordhuis: or do you want to test if callbacks are getting called?
09:48:07  <bnoordhuis>indutny: i want to test all code paths if possible
09:48:22  <indutny>oh
09:48:24  <indutny>even that :D
09:48:28  <indutny>I'll think about it
09:49:05  <indutny>fighting v8's debugger atm
09:52:34  * loladirojoined
09:54:36  <indutny>oh
09:54:40  <indutny>I think I've found a bug
09:54:44  <indutny>cra
09:54:45  <indutny>p
09:54:46  <indutny>:D
09:57:26  <indutny>bnoordhuis: why semwait may return EFAULT?
09:57:35  <bnoordhuis>indutny: what os?
09:57:42  * bradleymeckquit (Quit: bradleymeck)
09:58:00  <indutny>mac os
09:58:02  <indutny>semaphore_wait
09:58:40  <bnoordhuis>indutny: i don't know, you wrote it :)
09:59:06  <indutny>in v8
09:59:07  <indutny>:D
09:59:16  <bnoordhuis>where in v8?
09:59:26  <indutny>may be use after free?
09:59:31  <indutny>platform-macos.cc
09:59:44  * loladiroquit (Ping timeout: 240 seconds)
10:00:41  * AvianFlujoined
10:00:54  <indutny>looks like some memory corruption there
10:01:25  <bnoordhuis>indutny: well... the code doesn't check if semaphore_create() succeeds
10:01:34  <bnoordhuis>but could be memory corruption as well
10:02:27  <bnoordhuis>the linux implementation has the same bug...
10:02:52  <bnoordhuis>but on linux semaphores are implemented with futexes and those never fail
10:03:03  <indutny>destroy: 0xfc080
10:03:11  <indutny>this can't be good
10:03:16  <indutny>destructor was called for this address
10:03:48  <bnoordhuis>well, that explains the EFAULT :)
10:04:17  <indutny>haha
10:04:26  <indutny>probably segfault happens
10:04:31  <indutny>and semaphore_wait returns
10:04:35  <indutny>but sounds really odd
10:05:12  * AvianFluquit (Client Quit)
10:05:46  <indutny>ah
10:05:50  <indutny>this is not a errno
10:05:53  <indutny>this is KERN_ABORTED
10:07:15  <indutny>odd odd osx
10:07:30  <indutny>yeah, no segfaults
10:07:33  <indutny>gtg
10:07:39  <indutny>should be back in 1 hour or so
10:07:41  <indutny>(fixed)
10:20:39  * bradleymeckjoined
10:20:49  * bradleymeckquit (Client Quit)
10:34:41  * piscisaureus_joined
11:01:16  * piscisaureus__joined
11:01:16  * piscisaureus_quit (Read error: Connection reset by peer)
11:01:26  * piscisaureus__changed nick to piscisaureus_
11:04:21  * stagasjoined
11:12:17  * dylangjoined
11:19:44  * piscisaureus_quit (Ping timeout: 250 seconds)
11:24:20  * abraxasquit (Remote host closed the connection)
11:30:42  * dylangquit (Quit: dylang)
11:32:13  * dylangjoined
11:41:03  * piscisaureus_joined
11:44:31  * dylangquit (Quit: dylang)
11:51:42  <indutny>piscisaureus_: hey man
11:51:45  <indutny>I've fixed that problem
12:07:02  <indutny>piscisaureus_: https://chromiumcodereview.appspot.com/10867009
12:12:02  <indutny>bnoordhuis: btw, why ain't we processing SIGCHLD?
12:12:15  <bnoordhuis>indutny: ?
12:12:42  <indutny>I mean should we?
12:12:47  <indutny>there're some stuff with wait() call
12:12:53  <indutny>which I do not really understand so far
12:13:00  * mmaleckiquit (Quit: leaving)
12:13:07  <bnoordhuis>"some stuf""
12:13:39  <bnoordhuis>indutny: so what exactly is the issue?
12:14:01  <indutny> If a process explicitly specifies SIG_IGN as the action for the signal SIGCHLD, the system will not create zombie processes when children of the calling process exit. As a
12:14:01  <indutny> consequence, the system will discard the exit status from the child processes. If the calling process subsequently issues a call to wait(2) or equivalent, it will block until
12:14:01  <indutny> all of the calling process's children terminate, and then return a value of -1 with errno set to ECHILD.
12:14:08  <indutny>zombie processes
12:15:16  <bnoordhuis>i'll get my chainsaw and shotgun
12:15:35  <bnoordhuis>what wait() call are you talking about btw?
12:16:12  <indutny>I duno
12:16:15  <indutny>dunno*
12:16:22  <indutny>I just accidently found this description in my man
12:16:52  <indutny>looks like they're talking about `wait(NULL)` call
12:17:40  <bnoordhuis>indutny: yes. libuv doesn't use wait(), it uses waitpid(WNOHANG)
12:17:52  <indutny>ah, ok
12:18:11  <indutny>bnoordhuis: via libev?
12:18:38  <indutny>em...
12:18:44  <indutny>I don't think that we're using it
12:19:17  * piscisaureus_quit (Ping timeout: 244 seconds)
12:19:27  <indutny>ircretary: tell piscisaureus_ about https://chromiumcodereview.appspot.com/10867009/
12:19:27  <ircretary>indutny: I'll be sure to tell piscisaureus_
12:19:30  <indutny>thanks
12:21:27  <bnoordhuis>indutny: check uv__chld() in src/unix/process.c
12:24:47  <indutny>oh, it's only in master
12:27:23  <indutny>bnoordhuis: btw, are you accepting patches for uv?
12:29:26  <bnoordhuis>indutny: no, it's finished, done for all eternity
12:29:35  <bnoordhuis>what do you want to patch?
12:29:35  <indutny>oh
12:29:36  <indutny>ok
12:29:55  <indutny>bnoordhuis: https://github.com/joyent/libuv/pull/533
12:30:06  <indutny>uv_sem_wait
12:30:47  <indutny>brb
12:31:28  * piscisaureus_joined
12:31:42  <piscisaureus_>ircretary: notes
12:32:20  <bnoordhuis>indutny: expr != KERN_SUCCESS is kind of superfluous, KERN_SUCCESS == 0
12:33:15  <bnoordhuis>indutny: also, why loop when r == KERN_ABORTED? that doesn't look right
12:33:26  <piscisaureus_>indutny: this fixes the debugger crash
12:33:28  <piscisaureus_>?
12:47:22  <indutny>piscisaureus_: yes
12:47:32  <indutny>bnoordhuis: because KERN_ABORTED is returned when process receives signal
12:47:36  <indutny>sort of EINTR
12:47:37  <CIA-131>node: Bert Belder openshift * rcd23e6d / (10 files in 6 dirs): Make node work on Openshift - http://git.io/GM3IZQ
12:47:37  <CIA-131>node: Bert Belder openshift * r00cbcfc / (lib/dgram.js lib/net.js): openshift: listen on internal IP when no IP is specified - http://git.io/QqfiQw
12:47:38  <CIA-131>node: Bert Belder openshift * r5be1814 / (470 files in 62 dirs): Merge commit 'v0.6.21' into openshift - http://git.io/AQfZ4g
12:47:39  <CIA-131>node: Bert Belder openshift * rd7429dd / lib/readline.js : readline: make --interactive work when stdin is a pipe and stdout is a TTY - http://git.io/65dUOA
12:47:41  <CIA-131>node: Nathan Rajlich openshift * r2040ca9 / (src/node.cc src/node.js test/simple/test-force-repl.js): cmd: add a -i / --interactive flag to force the REPL - http://git.io/X1QXZQ
12:48:05  <piscisaureus_>ah
12:48:07  <piscisaureus_>sorry
12:48:08  <piscisaureus_>shift
12:48:11  <piscisaureus_>shit
12:48:22  <indutny>eh
12:48:29  <piscisaureus_>that was the wrong remote :-)
12:49:03  <piscisaureus_>I deleted the branch already
12:50:16  <indutny>oh
12:50:25  <indutny>I think there's a good git conf for this
12:50:35  <piscisaureus_>nah
12:50:35  <indutny>[push]
12:50:35  <indutny> default = tracking
12:50:46  <piscisaureus_>I always push with "git push origin HEAD:master"
12:51:13  <piscisaureus_>but I should've done "git push ajaxorg HEAD:openshift" here
12:51:30  <bnoordhuis>indutny: KERN_ABORTED is sent when another thread is killed, right? how do you know the semaphore is in a consistent state after that?
12:51:46  <indutny>bnoordhuis: nope, not only
12:51:48  <indutny>SIGCHLD
12:51:53  <indutny>will cause it to happen too
12:52:18  <bnoordhuis>really? mach is so odd
12:52:28  <indutny>http://qt.gitorious.org/qt/qtbase/commit/a6c8d0ba56d28f306a66f76c8aedd189003b1080?format=patch
12:52:36  <indutny>everyone is doing this
12:52:53  <indutny>http://golang.org/src/pkg/runtime/thread_darwin.c
12:52:55  <indutny>everyone
12:53:00  <indutny>except v8 and us
12:53:07  <indutny>and yes, mach is wierd
12:53:59  <bnoordhuis>so how do you know if it's SIGCHLD or thread termination?
12:54:20  <indutny>no way to know :P
12:54:36  <indutny>The caller was blocked due to a negative count on the semaphore, and was awoken for a reason not related to the semaphore subsystem (e.g. thread_terminate).
12:54:38  * dylangjoined
12:54:48  <indutny>that's what man says
12:55:22  <indutny>so it really means anything
12:55:48  <indutny>http://www.opensource.apple.com/source/Libc/Libc-391/pthreads/pthread_mutex.c
12:55:59  <indutny>bnoordhuis: apple is looping on it too
12:56:34  <bnoordhuis>okay
12:58:42  <indutny>bnoordhuis: fixed https://github.com/indutny/libuv/commit/4cbfd209f401fb03e45ade302091157a1ca33e1e
13:01:08  <bnoordhuis>indutny: btw, what about uv_sem_trywait?
13:02:11  <indutny>idk
13:02:15  <indutny>it's not really documented
13:02:32  <indutny>but it'll return some non-zero value
13:02:49  <indutny>anyway
13:02:59  <CIA-131>libuv: Fedor Indutny master * r8969df6 / src/unix/thread.c : darwin: handle KERN_ABORTED in uv_sem_wait() - http://git.io/q9lLRg
13:04:42  <CIA-131>libuv: Ben Noordhuis master * r120e2c1 / src/unix/thread.c : darwin: fix return value of uv_sem_init() - http://git.io/VwoAFw
13:04:51  * travis-cijoined
13:04:51  <travis-ci>[travis-ci] joyent/libuv#610 (master - 8969df6 : Fedor Indutny): The build passed.
13:04:51  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/29eb460d0b20...8969df6015a2
13:04:51  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2200351
13:04:51  * travis-cipart
13:05:43  <indutny>cool
13:05:54  <indutny>thank you
13:06:34  * travis-cijoined
13:06:34  <travis-ci>[travis-ci] joyent/libuv#611 (master - 120e2c1 : Ben Noordhuis): The build passed.
13:06:34  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/8969df6015a2...120e2c1335e9
13:06:34  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2200380
13:06:34  * travis-cipart
13:16:56  * stagasquit (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.0.17/2009122204])
13:24:53  * abraxasjoined
13:27:22  * piscisaureus__joined
13:29:04  * abraxasquit (Ping timeout: 240 seconds)
13:29:06  * piscisaureus_quit (Ping timeout: 260 seconds)
13:32:36  * piscisaureus__quit (Ping timeout: 260 seconds)
13:40:40  * loladirojoined
13:41:51  * loladiroquit (Client Quit)
13:50:06  * dylangquit (Quit: dylang)
14:03:54  * chobi_e_changed nick to chobi_e
14:16:48  * tomshredsjoined
14:28:42  * TheJHquit (Read error: Operation timed out)
14:28:47  * dylangjoined
14:36:10  * mmaleckijoined
14:39:41  <indutny>bnoordhuis: this mach thing scaries me much
14:39:47  <indutny>I can't reproduce it
14:43:43  * c4milojoined
14:45:24  * hzquit (Ping timeout: 272 seconds)
14:45:47  * hzjoined
14:47:49  <indutny>what can I receive when child process was killed with SIGTERM
14:47:53  <indutny>bnoordhuis: ^ ?
14:48:14  * loladirojoined
14:51:06  <indutny>hm...
14:51:10  <indutny>only SIGCHLD was emitted
14:52:38  * loladiroquit (Ping timeout: 245 seconds)
14:53:27  * loladirojoined
15:02:03  * loladiroquit (Quit: loladiro)
15:04:09  * loladirojoined
15:08:04  * dylangquit (Quit: dylang)
15:16:38  * AvianFlujoined
15:16:47  <indutny>ok, I got it
15:19:32  * piscisaureus_joined
15:26:42  <indutny>bnoordhuis: I'm afraid that we might made a bad thing there
15:26:49  <indutny>piscisaureus_: looks like I've overlooked this problem
15:26:56  <indutny>something strange is happening there
15:31:12  * chobi_echanged nick to chobi_e_
15:31:51  <piscisaureus_>indutny: huh. what?
15:32:00  <piscisaureus_>indutny: to me it seems you did the right thing
15:36:22  <indutny>nope
15:36:23  <indutny>I didn't
15:36:29  <indutny>things are more complicated there
15:36:52  <indutny>yeah
15:36:53  <indutny>just fixed
15:38:19  * TooTallNatejoined
15:38:23  <indutny>yeah
15:38:39  <indutny>void Debugger::NotifyMessageHandler(v8::DebugEvent event,
15:38:47  <indutny>looks like this thing was executed from two different threads
15:43:44  <indutny>piscisaureus_: console.log('past debugger statement');
15:43:47  <indutny>what is this^
15:43:57  <indutny>this should not be called, right?
15:44:13  <piscisaureus_>indutny: well, only after you "continue" :-)
15:44:18  <indutny>heh
15:44:24  <piscisaureus_>indutny: it also gets called when you don't debug
15:44:53  <indutny>ok, may be I was doing a right thing there
15:44:58  <indutny>it should hang
15:45:01  <indutny>and infinitely loop, right?
15:46:17  <indutny>yeah, it works fine
15:46:21  <indutny>hahaha
15:46:25  <indutny>so many temptations
15:46:34  <indutny>now I need to find a way to reproduce it
15:47:16  <indutny>I wonder why it's getting aborted
16:05:22  <indutny>write(0x5, "\0", 0x1) = 1 0
16:05:22  <indutny>sigreturn(0x7FFF5FBFE520, 0x1E, 0x0) = 0 Err#-2
16:05:28  <indutny>what can this mean?
16:05:35  * dapjoined
16:05:36  <indutny>especially fd=5
16:07:34  <TooTallNate>call?
16:07:57  * TheJHjoined
16:09:29  <indutny>huh?
16:09:36  <TooTallNate>sykpe?
16:09:44  <indutny>TooTallNate: Thursday
16:09:51  <indutny>call is on Thursday
16:09:52  <TooTallNate>oh wow
16:09:55  <indutny>usually :D
16:10:00  <TooTallNate>i thought it was thursday…
16:10:02  <TooTallNate>wtf
16:10:04  <tjfontaine>TooTallNate: shall we get you a calendar? :)
16:10:13  <TooTallNate>wow, how did that just happen, hahaha
16:10:36  <TooTallNate>guess that explains why my calendar notification *didn't* go off
16:10:52  <TooTallNate>ok, i'm off to the office then, back on in a bit
16:10:55  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
16:11:01  <indutny>ok
16:23:04  * ikpart
16:23:04  * ericktquit (Ping timeout: 252 seconds)
16:25:00  * joshthecoderjoined
16:30:35  <indutny>piscisaureus_: so for me KERN_ABORTED happens only if I'll call pthread_cancel
16:31:04  <indutny>so far I hadn't found another way to do it
16:31:12  <piscisaureus_>indutny: ok... so it doesn't happen when a signal is raised?
16:31:52  <piscisaureus_>indutny: can you share your test code
16:31:58  <indutny>yeah, sure
16:32:55  <indutny>https://gist.github.com/81c3620a29d9e55fcc72
16:33:05  <indutny>there're bunch of testing code
16:33:10  <indutny>which doesn't affect execution
16:33:13  <indutny>like second thread
16:33:20  <indutny>but you may try doing thread_suspend
16:33:24  <indutny>or thread_terminate from it
16:33:34  <piscisaureus_>indutny: so what if you add a signal handler
16:33:42  <piscisaureus_>indutny: nvm - I will try myself
16:33:44  * piscisaureus_reboots
16:33:47  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
16:39:14  <isaacs>good morning
16:39:28  <indutny>morning man
16:40:13  * arlolrajoined
16:45:56  * perezdjoined
16:47:33  <CIA-131>node: isaacs v0.8.8-release * r513a397 / lib/tls.js : lint - http://git.io/sKT1GQ
16:47:34  <CIA-131>node: isaacs v0.8.8-release * r0ddc566 / (AUTHORS ChangeLog src/node_version.h): wip - http://git.io/BTMY7w
17:02:11  <indutny>isaacs: what are main nits of upcoming release?
17:07:01  * piscisaureujoined
17:07:15  <piscisaureu>hey isaacs
17:07:25  <indutny>piscisaureu: hey man
17:07:27  <piscisaureu>isaacs: in node 0.8.7 npm is broken on windows
17:07:34  <indutny>piscisaureu: things are really complicated over there
17:07:38  <indutny>piscisaureu: I'm reading xnu source
17:07:40  <piscisaureu>isaacs: it installs global modules into the program files folder
17:11:47  <indutny>piscisaureu: also, thread_abort will cancel it
17:13:03  <indutny>and looks like this is the only reason
17:13:30  <isaacs>piscisaureu: testing
17:14:13  <indutny>1ooooh
17:14:15  <indutny>crap
17:14:51  * ericktjoined
17:17:40  <piscisaureu>indutny: I can reproduce it
17:18:05  <piscisaureu>indutny: a signal will interrupt semaphore_wait() *iff* you handle the signal
17:18:51  <indutny>oh
17:18:54  <indutny>oh noes
17:18:55  <indutny>:D
17:19:01  <indutny>this is definitely a bug
17:19:09  <indutny>I can show you a source that is causing it
17:19:28  * erickt_joined
17:19:31  <indutny>piscisaureu: https://gist.github.com/c31acdd03da6dd606d34
17:19:57  <piscisaureu>indutny: so if I install a SIGCHLD handler and then send it to a process it interrupts the wait
17:19:57  <piscisaureu>indutny: https://gist.github.com/3427691
17:19:57  <piscisaureu>indutny: try to send SIGUSR1 to it
17:20:06  <indutny>nice
17:20:13  <indutny>yeah, we're handling SIGUSR1
17:20:24  <indutny>but are we handling SIGCHLD anywhere?
17:20:31  <piscisaureu>indutny: well, of course
17:20:42  <piscisaureu>indutny: libev does that to detect exiting processes
17:21:03  <indutny>nvm
17:21:06  <piscisaureu>indutny: so if you kill a child process then SIGCHLD will be sent to node
17:21:13  <indutny>piscisaureu: btw, if you'll send SIGCHLD
17:21:15  <piscisaureu>indutny: and that will interrupt the semaphore
17:21:20  <indutny>will it interrupt wait?
17:21:21  * ericktquit (Ping timeout: 256 seconds)
17:21:21  * erickt_changed nick to erickt
17:21:24  <indutny>because it should not
17:21:28  <piscisaureu>indutny: lemme try
17:21:30  <indutny>as far as I can see from xnu's source
17:21:39  <indutny>only signals that should kill process will terminate call
17:21:42  <isaacs>piscisaureu: afaict, it works properly with npm 1.1.58
17:21:43  <indutny>as I can guess atm
17:21:48  <isaacs>piscisaureu: probably got fixed along the way :)
17:21:59  <piscisaureu>indutny: exactly the same behaviour
17:22:06  <piscisaureu>isaacs: are you sure?
17:22:15  * stagasjoined
17:22:19  <isaacs>https://gist.github.com/3427707
17:22:23  <isaacs>piscisaureu: ^
17:22:32  <isaacs>piscisaureu: how'd you install node/npm?
17:22:33  <isaacs>using the msi?
17:22:40  <piscisaureu>isaacs: Iya
17:22:44  <piscisaureu>isaacs: ya
17:22:54  <isaacs>piscisaureu: hm. what's `npm config ls` output?
17:23:03  <piscisaureu>isaacs: have to reboot. A sec
17:23:06  <isaacs>k
17:23:08  * piscisaureuquit (Quit: Leaving.)
17:23:13  * isaacsheh. windows.
17:23:17  <isaacs>"reboot"
17:23:34  <indutny>philips_: oddd
17:23:36  <indutny>oops
17:23:41  <indutny>bert just left
17:24:40  * piscisaureus_joined
17:25:36  * joshthecoderquit (Quit: Linkinus - http://linkinus.com)
17:25:39  <piscisaureus_>hello isaacs
17:25:48  <indutny>piscisaureus_: hey man
17:25:55  <indutny>piscisaureus_: I'm really surprised that this is happening
17:26:02  <isaacs>piscisaureus_: wait, my test was invald.
17:26:05  <isaacs>one se
17:26:06  <isaacs>c
17:26:07  * brsonjoined
17:26:16  <philips_>indutny: huh?
17:26:29  <philips_>indutny: oh you were trying to ping piscisaureus_
17:26:36  * TooTallNatejoined
17:27:05  <piscisaureus_>isaacs: https://gist.github.com/3427739
17:27:09  <indutny>aaah
17:27:11  <isaacs>piscisaureus_: yeah, i see it
17:27:30  <piscisaureus_>isaacs: 1.1.56 is pretty new :-)
17:28:04  <indutny>piscisaureus_: https://gist.github.com/2b7d513c2d406228acee
17:28:23  <indutny>piscisaureus_: else branch
17:33:56  * piscisaureus__joined
17:34:06  * chilts_joined
17:34:07  * einaros_joined
17:34:34  * russfran1joined
17:35:42  * piscisaureus___joined
17:37:02  * piscisaureus___quit (Read error: Connection reset by peer)
17:38:17  * piscisaureus__quit (Read error: Connection reset by peer)
17:38:51  * ryahquit (*.net *.split)
17:38:51  * russfrankquit (*.net *.split)
17:38:51  * chiltsquit (*.net *.split)
17:38:51  * einarosquit (*.net *.split)
17:38:53  * philips_quit (*.net *.split)
17:38:53  * deoxxaquit (*.net *.split)
17:38:53  * isaacsquit (*.net *.split)
17:38:53  * DrPizzaquit (*.net *.split)
17:38:53  * jcequit (*.net *.split)
17:38:53  * piscisaureus_quit (*.net *.split)
17:38:53  * perezdquit (*.net *.split)
17:38:54  * toothrquit (*.net *.split)
17:38:54  * bweaverquit (*.net *.split)
17:38:54  * AlbireoXquit (*.net *.split)
17:38:55  * zippquit (*.net *.split)
17:38:55  * russell_hquit (*.net *.split)
17:38:55  * ircretaryquit (*.net *.split)
17:38:55  * pquernaquit (*.net *.split)
17:38:55  * rphillipsquit (*.net *.split)
17:38:56  * brsonquit (*.net *.split)
17:38:56  * `3rdEdenquit (*.net *.split)
17:38:56  * mjr_quit (*.net *.split)
17:38:57  * indutnyquit (*.net *.split)
17:38:58  * voodootikigodquit (*.net *.split)
17:38:58  * chrisdickinsonquit (*.net *.split)
17:38:58  * tjfontainequit (*.net *.split)
17:38:58  * bnoordhuisquit (*.net *.split)
17:38:59  * rendarquit (*.net *.split)
17:38:59  * gicodequit (*.net *.split)
17:38:59  * Benviequit (*.net *.split)
17:39:00  * rvaggquit (*.net *.split)
17:39:20  * brsonjoined
17:39:20  * `3rdEdenjoined
17:39:20  * mjr_joined
17:39:44  * ryah_joined
17:39:44  * philips_joined
17:39:44  * deoxxajoined
17:39:44  * isaacsjoined
17:39:44  * DrPizzajoined
17:39:44  * jcejoined
17:39:56  * rendarjoined
17:39:56  * gicodejoined
17:39:56  * Benviejoined
17:39:56  * rvaggjoined
17:40:57  * indutnyjoined
17:40:57  * voodootikigodjoined
17:40:57  * chrisdickinsonjoined
17:40:57  * tjfontainejoined
17:41:00  * zippjoined
17:41:00  * russell_hjoined
17:41:00  * ircretaryjoined
17:41:00  * rphillipsjoined
17:41:00  * pquernajoined
17:41:08  * stephankjoined
17:41:43  * perezdjoined
17:41:43  * toothrjoined
17:41:43  * bweaverjoined
17:41:43  * AlbireoXjoined
17:41:52  * piscisaureus_joined
17:42:01  <piscisaureus_>hmm my messages never arrived
17:42:02  <piscisaureus_>indutny: ok, here's the news
17:42:02  <piscisaureus_>indutny: if the kernel needs to run user code, it always makes a thread runnable
17:42:02  <piscisaureus_>indutny: which one it is isn't defined (but in a single threaded test that's pretty obvious heh :-))
17:42:02  <piscisaureus_>indutny: so, *unless* semaphore_wait restarts the syscall after that happens, a handled signal will always cause it to return.
17:42:02  <piscisaureus_>indutny: so you should probably take a look at the semaphore_wait source code
17:42:02  <piscisaureus_>indutny: because the signal stuff won't tell you anything new
17:43:13  <indutny>piscisaureus_: well, I looked at it
17:43:15  <indutny>and over it
17:43:15  <indutny>:)
17:44:10  <indutny>but I think it may be handled internally
17:44:18  <indutny>or
17:44:21  <indutny>at least documented well
17:45:03  * erickt_joined
17:45:55  * brsonquit (*.net *.split)
17:45:55  * `3rdEdenquit (*.net *.split)
17:45:55  * mjr_quit (*.net *.split)
17:46:16  * brsonjoined
17:46:16  * `3rdEdenjoined
17:46:16  * mjr_joined
17:46:58  <piscisaureus_>indutny: anyway, my test shows that it can be interrupted right :-)
17:47:03  <indutny>haha
17:47:04  <indutny>yes
17:47:11  <indutny>this is because thread_abort is called internally
17:47:17  <indutny>from a process signal handler
17:47:22  <indutny>which is inside xnu ;)
17:47:29  <indutny>have you seen my gist?
17:47:53  * ericktquit (Ping timeout: 252 seconds)
17:47:53  * erickt_changed nick to erickt
17:47:54  * paddybyersjoined
17:48:13  <piscisaureus_>indutny: I did - but no threads actually get aborted
17:48:19  <indutny>no, they're
17:48:24  <indutny>it's called thread_abort_safely
17:48:30  <indutny>and what it actually does
17:48:39  <indutny>is leaving syscall and returning that stupid error
17:48:39  <indutny>:D
17:53:56  <CIA-131>node: isaacs v0.8.8-release * r3c89964 / (102 files in 7 dirs): npm: Upgrade to 1.1.59 - http://git.io/t9xo5Q
17:53:56  <CIA-131>node: isaacs v0.8.8-release * r2fe8613 / (AUTHORS ChangeLog src/node_version.h): wip - http://git.io/3sXPsQ
17:53:59  <isaacs>piscisaureus_: can you test with that? ^
17:54:13  <isaacs>piscisaureus_: i'd had a global config that was overriding my lack of builtinconfig being respected
17:55:01  <pquerna>indutny: i just read your tls sessionc ache patch; crazy approach :P
17:57:04  <piscisaureus_>isaacs: ure
17:57:18  <isaacs>TooTallNate: you updated node-gyp *again*?
17:57:27  * isaacs<-- hypocrite ;)
17:57:32  <TooTallNate>isaacs: :p
17:57:37  <TooTallNate>ya another small patch
17:57:42  <isaacs>k
17:57:49  <isaacs>can it wait? or is it a bug fix worth pulling in?
17:57:54  <isaacs>this is getting a little bit ridiculous...
17:57:58  <TooTallNate>it can wait
17:58:01  <isaacs>kewl, thanks :)
17:58:10  <TooTallNate>hahah
17:58:16  <TooTallNate>like 3 npm updates in one release?
18:02:33  <indutny>pquerna: haha
18:02:35  <indutny>pquerna: it works
18:03:24  <piscisaureus_>isaacs: yep, works
18:03:42  <piscisaureus_>isaacs: 1.1.48 -> 1.1.59 ?
18:04:36  <CIA-131>node: isaacs v0.8.8-release * ra299c97 / (AUTHORS ChangeLog src/node_version.h): 2012.08.22, Version 0.8.8 (Stable) - http://git.io/rFUufQ
18:04:37  <CIA-131>node: isaacs v0.8.8-release * ree20094 / lib/tls.js : lint - http://git.io/Yr_gsA
18:04:37  <CIA-131>node: isaacs v0.8.8-release * rc259499 / (103 files in 7 dirs): npm: Upgrade to 1.1.59 - http://git.io/T0DONA
18:04:55  <isaacs>piscisaureus_: there were 11 bugfixes :)
18:06:28  * lohkeyjoined
18:06:47  <isaacs>piscisaureus_: part of the issue was that i use npm in the npm website, and i was deploying each fix to it
18:06:56  <isaacs>usually i don't actually do as many npm releases
18:07:03  <isaacs>it's fine. it's just a number :)
18:07:47  * isaacstopic: Liberty. Unity. Velocity. http://static.izs.me/irclogs/libuv/
18:14:09  * hzquit
18:15:18  * joshthecoderjoined
18:17:03  * bnoordhuis_joined
18:17:03  topic: v0.8.8 release tomorrow
18:18:53  <ryah_>~is there going to be a new v0.9 ?
18:20:19  * ericktquit (Quit: erickt)
18:22:26  <indutny>ryah_: what do you mean? 0.9.0 release?
18:26:09  <indutny>ryah_: btw, have you tried vock? ;)
18:26:47  <piscisaureus_>indutny: good work on the v8 patch btw
18:26:58  <indutny>piscisaureus_: not really, you did a lot
18:27:04  <indutny>but I was getting close to it anyway :D
18:30:33  * loladiroquit (Quit: loladiro)
18:34:01  * mikealquit (Quit: Leaving.)
18:34:46  <piscisaureus_>bnoordhuis_: hey, yt>
18:41:59  <CIA-131>libuv: Bert Belder v0.8 * ra787a16 / src/unix/async.c : unix: fix uv_async_send not working with Sun Studio (+108 more commits...) - http://git.io/QcM_Ow
18:43:12  <CIA-131>libuv: Bert Belder master * ra787a16 / src/unix/async.c : unix: fix uv_async_send not working with Sun Studio - http://git.io/QcM_Ow
18:43:49  <piscisaureus_>hmm
18:43:52  <piscisaureus_>not good
18:43:59  * travis-cijoined
18:44:00  <travis-ci>[travis-ci] joyent/libuv#612 (v0.8 - a787a16 : Bert Belder): The build was fixed.
18:44:00  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/a7b83e0b9822...a787a16ac371
18:44:00  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2203942
18:44:00  * travis-cipart
18:44:05  <mmalecki>piscisaureus_: 108 more commits?
18:44:11  <piscisaureus_>ya, not good
18:45:01  * travis-cijoined
18:45:02  <travis-ci>[travis-ci] joyent/libuv#613 (master - a787a16 : Bert Belder): The build passed.
18:45:02  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/120e2c1335e9...a787a16ac371
18:45:02  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2204009
18:45:02  * travis-cipart
18:48:17  <indutny>108 commits?!
18:48:23  <indutny>piscisaureus_: man, I'm going to kill you
18:48:31  <indutny>sorry
18:48:56  <TooTallNate>i think CIA was just being dumb
18:49:10  <piscisaureus_>I might have merged master accidentally
18:49:55  <tjfontaine>yes it seems like it
18:51:02  * isaacstopic: Liberty. Unity. Velocity. http://static.izs.me/irclogs/libuv/
18:51:10  <isaacs>wtf, pratchett?
18:51:42  <isaacs>ryah_: yeah, we'll do a 0.9.1 soon
18:52:01  <piscisaureus_>indutny: TooTallNate: mmalecki: fixed
18:52:42  <indutny>nice
18:54:51  * mikealjoined
18:56:13  * piscisaureus__joined
18:58:30  * piscisaureus_quit (Read error: Operation timed out)
19:00:25  <isaacs>tests all pass on os x and sunos
19:00:29  <isaacs>waiting on the linux brothers
19:02:25  <ryah_>isaacs: is npmjs.org going to get docs still?
19:02:29  <ryah_>i love the website
19:02:31  <ryah_>it's super fast
19:03:11  * piscisaureus___joined
19:03:16  * piscisaureus___changed nick to piscisaureus_
19:03:43  <isaacs>ryah_: thanks!
19:03:59  <isaacs>ryah_: so, i've been trying to find time to really sketch out the best plan for this.
19:04:10  * piscisaureus__quit (Ping timeout: 246 seconds)
19:04:16  <ryah_>it's fucking amazingly fast
19:04:21  <isaacs>obviously we need much better integration between userland modules, npm docs, and node docs
19:04:24  <ryah_>is it all statically generated?
19:04:28  <isaacs>ryah_: nope.
19:04:32  <isaacs>ryah_: it's mostly templates.
19:04:42  <ryah_>damn
19:04:44  <isaacs>the static files use my st module, which cacehs the shit out of every single thign it can.
19:05:02  <isaacs>and templar is very smart about caching templates based on the signature of the data object and the template fileame
19:05:04  <ryah_>maybe im just really close to the server
19:05:19  <ryah_>15ms
19:05:23  <ryah_>that doesn't hurt
19:05:35  <isaacs>so if your request ends up trying to generate the same template with the saem data, you get the cached contents, or usually a 304
19:06:02  <isaacs>ryah_: there are still some random cases where something internal in cluster just freaks the fuck out, and a req won't be responded to for 30 seconds until i time it out
19:06:07  * einaros_changed nick to einaros
19:06:08  <isaacs>ryah_: that's not so fast :)
19:06:11  <isaacs>but that's less common now.
19:06:24  <ryah_>hm
19:06:32  <ryah_>and it's all https?
19:06:36  <ryah_>daaang
19:06:36  <isaacs>yeah
19:06:42  <ryah_>this is served from node?
19:06:46  <isaacs>yes.
19:06:47  <isaacs>:)
19:06:53  <isaacs>100% node.
19:07:07  <ryah_>:)
19:07:15  <ryah_>makes me happy
19:07:27  * piscisaureus_quit (Client Quit)
19:07:32  <ryah_>i like the tilda username
19:07:35  <ryah_>very cute
19:07:42  <isaacs>ryah_: yeah, that was konobi's idea
19:07:49  <isaacs>i wasnt' very happy with /profile/username
19:08:04  <isaacs>but /pkgname already redirects if it's not already something else
19:09:03  <ryah_>are the caches on dist?
19:09:07  <ryah_>*disk
19:09:23  <isaacs>ryah_: the data lives in: couchdb, disk, memory, your browser
19:09:28  <isaacs>and i try to cache as far to the right as possible
19:09:38  <isaacs>ryah_: i'm using lru-cache pretty heavily
19:11:49  <ryah_>isaacs: if you ever hear about someone wanting to do ninja tell them to do this issue: https://github.com/joyent/node/issues/3903
19:11:53  <ryah_>cause i want to use it
19:12:00  <isaacs>ryah_: yeah, it sounds neat
19:12:06  <isaacs>i +1 that issue :)
19:12:26  <isaacs>lol "i forgot how that works. i can wait until some intrepid soul does it :)
19:12:47  <isaacs>TooTallNate: want ryah_ to owe you a favor?
19:12:50  <isaacs>;P
19:13:15  <ryah_>oh yeah, TooTallNate is the gyp dude
19:13:26  <ryah_>TooTallNate: all the ninja stuff is already in there
19:13:35  <ryah_>just needs some massaging
19:13:50  <ryah_>(i actually tried and hit a snag and gave up)
19:14:08  <ryah_>btw are we keeping the GYP inside of Node current?
19:14:19  <TooTallNate>i'll mess around with it
19:14:24  <TooTallNate>i've never used ninja before though
19:14:37  <tjfontaine>it's pretty nice
19:14:44  <ryah_>ninja is made by my friend evan martin
19:14:53  <ryah_>who is the greatest programmer i know
19:15:00  <ryah_>and a major influence on me
19:15:28  <isaacs>tjfontaine: you could do it, too
19:15:32  <ryah_>so - although i haven't used ninja beyond a simple test - i imagine it's very very cool
19:15:46  <ryah_>chrome is using it now too - so it's far from a trivial little thing
19:15:58  <isaacs>yep
19:16:20  <tjfontaine>isaacs: I already am
19:16:30  <tjfontaine>or at least trying it out
19:16:49  * mikealquit (Quit: Leaving.)
19:17:56  <isaacs>kewl :)
19:21:17  * mikealjoined
19:23:10  <ryah_>so nice that there is no more waf
19:24:06  <ryah_>isaacs, TooTallNate: https://gist.github.com/87c7c21786ebc17d5a9f
19:24:48  <ryah_>tested on linux - probably should be tested on mac and windows too
19:24:57  <isaacs>ryah_: yeah, i'm a little concerned about the fallout about no waf
19:25:52  <isaacs>ryah_: thanks. TooTallNate: you wanna review that? if you say it works, i'll believe you :)
19:27:15  * AvianFluquit (Quit: AvianFlu)
19:29:19  <isaacs>ryah_: did you see the download button change?
19:30:28  <isaacs>ryah_: http://nodejs.org/docs/v0.8.8/ just downloads teh tarball on linux, which is kinda crappy, but deb/rpm is coming soonish
19:30:41  <isaacs>on windows/osx it downloas the installer
19:30:51  <isaacs>and picks the right x64/x86 one for windows
19:31:18  <ryah_>*shrug* what was wrong with the popup div
19:31:20  <ryah_>i liked that
19:31:37  <isaacs>ryah_: i grew to hate it
19:32:07  <isaacs>it's twice as many clicks as it should take to install something, breaks the back button in subtle ways, and obscures the options if you actually want to see the options
19:32:24  <isaacs>now we have a page with the options, and a button that installs in one click.
19:32:30  * mmaleckiquit (Ping timeout: 268 seconds)
19:32:34  <isaacs>and the back button works like a back button
19:34:10  <ryah_>*shrug*
19:35:07  <isaacs>when we get rpm/deb installers, it'll be pretty much ideal.
19:37:27  * bnoordhuis_is back
19:37:59  <ryah_>bnoordhuis_: is it okay to upgrade gyp? i seem to remember you floating some patches?
19:38:19  <ryah_>bnoordhuis_: https://gist.github.com/87c7c21786ebc17d5a9f
19:38:38  <bnoordhuis_>ryah_: it's okay, the bug's been fixed upstream
19:39:13  <bnoordhuis_>land at will
19:40:01  <CIA-131>node: Ryan Dahl v0.8 * rf90c9ce / (24 files in 5 dirs): Upgrade GYP to r1477 - http://git.io/o9G0hQ
19:41:43  <tjfontaine>oh that went into v0.8 not master
19:41:55  <ryah_>is that a problem?
19:42:18  <tjfontaine>I dunno, I did the small configure thing on master
19:42:40  <isaacs>it doesn't change any node api, though, does it?
19:42:46  <tjfontaine>no
19:43:03  <isaacs>yeah, i think v0.8 is fine, then
19:43:08  <ryah_>nod
19:43:39  <ryah_>tjfontaine: when you have that send it to me so i can test it
19:43:48  <ryah_>got ninja all ready to go
19:44:12  <isaacs>we'll have to make sure it doesn't break node-gyp, but that's something we can do later
19:44:22  <isaacs>just before 0.8.9
19:44:29  <tjfontaine>ryah_: short and sweet https://gist.github.com/3428699
19:44:32  <isaacs>and i'm not very nervous about it
19:45:31  <isaacs>we can land --with-ninja in 0.8 as well, as long as it's not default of course
19:45:50  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
19:46:14  <ryah_>tjfontaine: it would be nice if you could do
19:46:19  <ryah_>./configure --with-ninja
19:46:20  <ryah_>ninja
19:46:37  <ryah_>or - altrenatively since our makefile contains a bunch of node-specific things
19:46:38  <tjfontaine>ya, I figured that would be desirable, or perhaps a simple Makefile wrapper
19:46:42  <ryah_>./configure --with-ninja
19:46:43  <ryah_>make
19:46:47  <ryah_>^-- spawns ninja
19:46:50  <tjfontaine>nod
19:47:11  <ryah_>but i think just typing "ninja" what most people will want to do
19:47:45  <tjfontaine>I'll just make the decision of what to symlink based on if they passed debug
19:48:10  <tjfontaine>well, I guess ninja needs to cd
19:48:17  <tjfontaine>I'll figure it out
19:49:11  <ryah_>or maybe configure can just print "cd out/Release; ninja"
19:49:17  <ryah_>to tell people what to do
19:49:54  <tjfontaine>nod
19:50:23  <ryah_>i like how ninja shows the progress
19:52:24  <ryah_>tjfontaine: also think about how to do "make install" with ninja?
19:52:42  <tjfontaine>ryah_: and make test etc
19:54:37  <CIA-131>node: isaacs v0.8 * r0407ecf / doc/blog/release/v0.8.8.md : blog: v0.8.8 release (+5 more commits...) - http://git.io/FUYC7w
19:55:27  * chilts_changed nick to chilts
19:56:32  * AvianFlujoined
20:02:05  * txdvquit (Read error: Connection reset by peer)
20:03:51  * txdvjoined
20:11:38  * ericktjoined
20:16:35  * piscisaureus_joined
20:18:32  * `3rdEdenjoined
20:28:07  <bnoordhuis_>piscisaureus_: bertje, you rang?
20:28:25  <piscisaureus_>bnoordhuis_: yeah, but you answer 5 hours late
20:28:52  <bnoordhuis_>piscisaureus_: you in a rush?
20:28:59  <piscisaureus_>bnoordhuis_: no more.
20:29:09  <bnoordhuis_>good
20:29:17  <bnoordhuis_>stress is bad for one's health
20:31:17  <tjfontaine>ryah_: I've updated https://gist.github.com/3428699 wanna give that a try?
20:31:24  <ryah_>isaacs: http://nodejs.org/ looks a little messed up
20:31:32  <tjfontaine>ryah_: you should be able to configure --with-ninja && make
20:31:49  <isaacs>ryah_: oh, wild..
20:31:57  <piscisaureus_>oh
20:32:03  <bnoordhuis_>i like the minimal look
20:32:10  <piscisaureus_>isaacs: http://nodejs.org/dist/v0.8.8/node-v0.8.8-xx64.msi <-- the url is wrong
20:32:10  <isaacs>the industrygen script musta borked
20:32:27  <ryah_>tjfontaine: fatal: corrupt patch at line 90
20:32:47  <tjfontaine>heh, add a empty line :)
20:33:07  * stagas_joined
20:33:26  <ryah_>still broken
20:33:31  <tjfontaine>moment
20:34:30  <tjfontaine>I've updated again, is it still broken?
20:35:06  <ryah_>works
20:35:12  <tjfontaine>ok odd, but ok
20:35:30  <ryah_>tjfontaine: maybe --ninja instead of --with-ninja ?
20:35:37  * stagasquit (Ping timeout: 272 seconds)
20:35:44  * mmaleckijoined
20:35:47  * stagas_changed nick to stagas
20:35:54  <tjfontaine>since it's not a feature of node itself?
20:35:56  <isaacs>piscisaureus_: where do you see that url?
20:36:01  <ryah_>tjfontaine: yeah
20:36:01  <isaacs>piscisaureus_: the download button, or somwewhere else?
20:36:05  <piscisaureus_>isaacs: download button
20:36:16  <tjfontaine>ryah_: ya, ok
20:36:18  <ryah_> --with-something sounds like a feature of the build
20:36:32  <isaacs>ah, found it
20:36:34  <isaacs>fixed
20:36:53  <isaacs>piscisaureus_: reload?
20:37:11  <piscisaureus_>isaacs: http://nodejs.org/dist/v0.8.8/node-v0.8.8-x64.msi but still 404
20:37:18  <ryah_>i dont know about all these buttons on the home page
20:37:22  <ryah_>*shrug*
20:37:37  <isaacs>piscisaureus_: oh, right. needs x64/
20:37:45  <ryah_>maybe its okay
20:37:48  <piscisaureus_>I agree with ryah, it didn't really get much prettier
20:38:10  <isaacs>the two bottom buttons should be links
20:39:07  <piscisaureus_>I do like the 1-click download button btw
20:39:11  <isaacs>piscisaureus_, ryah_: patches welcome :)
20:39:29  <piscisaureus_>speaking of looks
20:39:39  <piscisaureus_>the new download page isn't very pretty either :-)
20:39:42  <isaacs>like most designers, i'm a terrible designer, but unlike most designers, i know it.
20:40:03  <isaacs>piscisaureus_: pretty it up,please.
20:40:09  <isaacs>piscisaureus_: it's just css and html and stuff.
20:40:15  <piscisaureus_>man
20:40:18  <piscisaureus_>that's years ago
20:40:25  <piscisaureus_>I must go back to w3schools first
20:40:28  <isaacs>hahahah
20:40:49  <piscisaureus_>isaacs: ok
20:41:21  <isaacs>piscisaureus_: the goal of the download button is for no one to have to see the download page.
20:41:53  <piscisaureus_>isaacs: as I said, I really like this download button
20:42:00  <isaacs>yeah
20:42:19  <isaacs>i don't know if there's a way to make the download page much much better, though
20:42:35  <isaacs>it's rather annoying when projects give you JUST a download button, but no way to download just the source, or a different arch or whatever.
20:42:46  <isaacs>sometimes i want to download the deb on my mac and scp it somewhere, or whatever.
20:42:47  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
20:42:59  * stephankchanged nick to Guest79945
20:43:36  * Guest79945quit (Quit: *Poof!*)
20:44:13  * stephankjoined
20:47:21  <piscisaureus_>isaacs: hey, you changed it already
20:48:37  <piscisaureus_>isaacs: I think we can just move the download button to the bottom-right menu
20:48:43  <piscisaureus_>isaacs: lemme fix that up
20:53:34  <piscisaureus_>I also wonder when the icons at the bottom right section became so random...
20:53:39  <piscisaureus_>although I never really notice
20:59:51  <piscisaureus_>let's delete the nodejs-dev mailing list
20:59:53  * AvianFluquit (Quit: AvianFlu)
21:01:51  <piscisaureus_>^-- isaacs
21:10:51  * tomshredsquit (Quit: Leaving...)
21:10:57  * mmaleckiquit (Quit: Reconnecting)
21:11:13  * mmaleckijoined
21:19:46  * dshaw_joined
21:20:03  * ericktquit (Quit: erickt)
21:20:43  <piscisaureus_>maybe not delete - but make it readonly
21:29:32  <bnoordhuis_>shut it down
21:29:40  <bnoordhuis_>i second this motion
21:37:56  * arlolraquit (Quit: Leaving...)
21:39:07  * stagasquit (Read error: Connection reset by peer)
21:41:37  * TooTallNatequit (Ping timeout: 265 seconds)
21:42:43  * TooTallNatejoined
21:46:13  <isaacs>piscisaureus_: icons at the bottom right section?
21:46:30  <piscisaureus_>isaacs: the menu with "about, docs, npm, community" etc
21:46:45  <piscisaureus_>isaacs: the docs have the community icon
21:46:53  <isaacs>oh, that
21:46:55  <piscisaureus_>isaacs: npm has the docs icon
21:46:56  <piscisaureus_>etc
21:46:56  <isaacs>yeah, that's crappy
21:47:17  <piscisaureus_>isaacs: there is this checkbox thing which I have no clue what it is for
21:47:22  <isaacs>piscisaureus_: no clue
21:47:30  <isaacs>piscisaureus_: i'm so tired of this website.
21:47:37  <piscisaureus_>isaacs: so ... tomorrow at the office (where I have photoshop) I will fix this
21:47:59  <piscisaureus_>isaacs: since you're out of energy :-)
21:48:10  <isaacs>piscisaureus_: it'd probably be best if the homepage had the download stuff, and then everythign below that was just like a "normal" page.
21:48:20  <isaacs>and move the "in the industry" to its own page.
21:48:32  <isaacs>and maybe ditch the icons in the siderail
21:48:34  <piscisaureus_>isaacs: I like the homepage somewhat
21:48:45  <isaacs>and like, that "featured" thing is 2 years old now.
21:48:46  <isaacs>srsly.
21:48:47  <piscisaureus_>isaacs: the "in the industry" section is good PR
21:48:51  <piscisaureus_>yeah
21:49:00  <isaacs>piscisaureus_: fine, but there's a bunch of them now, and we only ever show 4
21:49:07  <isaacs>it'd be nice if it showed one, and had a link to view wmore.
21:49:25  <piscisaureus_>isaacs: oh - I could consider that
21:49:35  <isaacs>and the "featured" link goes to a video that doesn't match the screenshot
21:49:37  <piscisaureus_>isaacs: it would also be nice if all the other pages where just white
21:49:43  <piscisaureus_>instead of brown
21:49:45  <isaacs>yeah
21:49:49  <isaacs>the brown is hard to look at.
21:49:58  <isaacs>basically, i think the docs are the correct style
21:50:01  <isaacs>the apidocs
21:50:24  <piscisaureus_>isaacs: the api docs look good
21:50:32  <isaacs>some of the code highlighting is a bit too light for the background, but that's very easily fixed.
21:50:45  <piscisaureus_>(although I miss an "index" a little)
21:51:09  <piscisaureus_>this is all very fixable
21:51:12  <isaacs>piscisaureus_: miss an "index"?
21:51:23  <isaacs>piscisaureus_: like, an alphabetized list of everything somewhere?
21:51:31  <piscisaureus_>isaacs: well, more like, better navigation
21:51:47  <piscisaureus_>isaacs: when I jump to e.g. fs.read then after that it's hard to get back
21:51:53  * arlolrajoined
21:51:54  <isaacs>gorsh, the "recommended modules" page... have you seen this lately?
21:51:55  * mmaleckiquit (Quit: Reconnecting)
21:52:07  * arlolraquit (Client Quit)
21:52:10  * mmaleckijoined
21:52:15  <piscisaureus_>no
21:52:18  <piscisaureus_>where's that?
21:52:19  <piscisaureus_>wiki?
21:52:21  <isaacs>http://nodejs.org/api/appendix_1.html
21:52:37  <piscisaureus_>yeah
21:52:39  <piscisaureus_>let's throw it out
21:52:54  <isaacs>+1
21:52:55  <isaacs>doit
21:52:55  <piscisaureus_>There are many third party modules for Node. At the time of writing, August 2010, the master repository of modules is the wiki page.
21:52:58  <piscisaureus_>^-- :-)
21:53:20  <isaacs>lol
21:53:42  <piscisaureus_>isaacs: anyway - regardless of how much we all hate it, I think the website is important
21:53:45  <piscisaureus_>like Windows
21:53:48  <isaacs>yes
21:53:57  <isaacs>that's why i haven't gone gangbusters on it too hard yet.
21:54:06  <isaacs>just made these stupid little incremental changes.
21:54:17  <isaacs>if you're going to do major surgery like that, you have to have some kind of plan, i think.
21:55:20  <piscisaureus_>isaacs: sure... I am not going to do major surgery
21:55:43  <piscisaureus_>isaacs: but I think I can contribute some incremental improvements
21:56:01  <isaacs>kewl :)
21:56:25  <isaacs>if you're going to update the link icons, wanna also retina-ize them?
21:58:07  <piscisaureus_>isaacs: ehm, maybe :-)
21:58:16  <piscisaureus_>isaacs: I don't know if my skill set allows that
21:58:34  <piscisaureus_>isaacs: you should know how much time this windows icon cost in the end :-)
21:58:54  <piscisaureus_>isaacs: remove the appendix from 0.8 ? or master only ?
21:58:57  <piscisaureus_>+1 for 0.8
21:59:00  * dapquit (Remote host closed the connection)
21:59:54  <TooTallNate>piscisaureus_: you just have to double the size basically
22:00:03  <TooTallNate>we did it for the "downloads" page for the platform icons
22:00:09  <TooTallNate>only 1 (large) size
22:00:16  <TooTallNate>and let the browser down-size it on regular displays
22:00:22  <piscisaureus_>TooTallNate: yeah, so I have to make the image sharper or something
22:00:48  <piscisaureus_>TooTallNate: unfortunately afaik background-image images don't scale
22:00:49  <TooTallNate>damn, don't we have svg versions anywhere or something?
22:00:55  <piscisaureus_>I don't
22:01:08  <TooTallNate>piscisaureus_: there's background-size, but it's somewhat newish
22:01:30  <piscisaureus_>so I'd have to convert it to <img> tags
22:01:33  <piscisaureus_>which is possible
22:02:06  <TooTallNate>piscisaureus_: is it safe to have the x86 and x64 node both installed at the same time?
22:02:11  <TooTallNate>using the msi
22:02:19  <piscisaureus_>TooTallNate: it's not possible atm
22:02:38  <isaacs>piscisaureus_: 0.8
22:02:46  <TooTallNate>is it cause it always installs to Progam Files (x86) or did that get fixed?
22:02:54  <CIA-131>node: Bert Belder v0.8 * r99ad52e / (3 files): doc: remote the "recommended modules" section - http://git.io/m5FD2w
22:03:05  <piscisaureus_>TooTallNate: no, it's because the msi uninstalls the other version
22:03:11  * dapjoined
22:03:13  <piscisaureus_>TooTallNate: although, it doesn't actually work
22:03:18  <TooTallNate>haha
22:03:20  <TooTallNate>damn installers
22:03:34  <piscisaureus_>TooTallNate: because of a WiX bug/quirk, the x32 installer can't uninstall the x64 version
22:03:39  <piscisaureus_>but vice versa it works
22:03:41  <piscisaureus_>:-(
22:03:49  <piscisaureus_>unless we use WiX 3.6-unstable
22:04:06  <piscisaureus_>TooTallNate: but I think I will change the package so you can have them both
22:04:23  <TooTallNate>then there's that PATH issue :p
22:04:33  <piscisaureus_>TooTallNate: yeah, that I cannot reproduce
22:04:53  <TooTallNate>i mean that both exe's are just called "node"
22:04:59  <TooTallNate>which issue you talking about?
22:05:01  * ericktjoined
22:05:28  <piscisaureus_>TooTallNate: the downside of allowing side by side installs is that there will be a "migration" where two versions of node can be installed together (like 0.8.8 and 0.8.9)
22:05:56  <TooTallNate>ya that's usually undesirable
22:06:18  <piscisaureus_>TooTallNate: https://github.com/joyent/node/issues/3902
22:06:34  <TooTallNate>oh right
22:06:36  <piscisaureus_>TooTallNate: renaming node.exe breaks compiled addons
22:07:11  <TooTallNate>piscisaureus_: ya the exe name is minor. it's not like python is called something else for the 64-bit version
22:07:12  <piscisaureus_>TooTallNate: there is a way around this - which involves delay loading hackery
22:12:48  <CIA-131>node: isaacs v0.8 * r1597ce0 / doc/index.html : website: Fix download link on win, un-break industry section - http://git.io/sNgkAQ
22:16:18  <TooTallNate>piscisaureus_: re node.exe name, have you heard any complaint about it? I haven't
22:16:34  <piscisaureus_>TooTallNate: I got a couple when node 0.6 was very young
22:17:12  <piscisaureus_>TooTallNate: but not recently no.
22:20:13  <TooTallNate>how do I debug this? https://gist.github.com/0b239712af1ea3abcebe
22:20:15  <TooTallNate>gdb?
22:21:30  <mmalecki>TooTallNate: I'd set up breakpoints on malloc and free
22:21:46  <mmalecki>but that's like shitload of breaks
22:22:12  <mmalecki>did you get a core?
22:22:19  <mmalecki>if so just `bt`
22:22:21  <mmalecki>in gdb
22:23:12  <TooTallNate>mmalecki: how do I generate a core again?
22:23:22  <mmalecki>TooTallNate: ulimit -c unlimited
22:23:26  <mmalecki>then run the app
22:23:28  <mmalecki>in the same shell
22:24:02  <TooTallNate>ok got one
22:24:08  <TooTallNate>how do i invoke gdb?
22:24:20  <mmalecki>TooTallNate: gdb <core-file>
22:24:52  <TooTallNate>weird
22:24:57  <TooTallNate>it says: ""/home/pi/node/core": not in executable format: File format not recognized"
22:25:34  <mmalecki>actually, wait
22:25:41  <mmalecki>try `gdb -c core`
22:25:53  <mmalecki>also I might or might notbe drunk
22:26:07  <TooTallNate>damn, i think i need a node_g
22:26:23  <TooTallNate>good 'ol cross compile!
22:26:26  <tjfontaine>gdb $(which node) core
22:26:46  <mmalecki>tjfontaine: ++
22:26:53  <mmalecki>wtf kohai
22:27:24  <tjfontaine>TooTallNate: iow it's gdb <exe> <core>
22:27:48  <TooTallNate>tjfontaine: trying
22:27:51  <tjfontaine>k
22:29:17  <TooTallNate>not much i think https://gist.github.com/17242120a9a5747b1a8c
22:29:34  <tjfontaine>what version of node is this?
22:30:01  <tjfontaine>I specifically remember bnoordhuis_ fixing this recently
22:30:48  <tjfontaine>a debug build would certainly be ideal for this in any event
22:31:28  <TooTallNate>tjfontaine: i remember a fix recently too
22:31:34  <TooTallNate>i thought it was for solaris though...
22:31:46  <tjfontaine>I'm not sure if it made it from libuv to node yet though
22:31:51  <tjfontaine>TooTallNate: there was another as well
22:31:52  <TooTallNate>it's some v0.8.8-pre btw
22:32:00  <TooTallNate>i'm gonna try the v0.8.8 tag
22:33:05  <TooTallNate><3 cross compiling
22:33:07  <tjfontaine>https://github.com/joyent/libuv/commit/54bfb66806a3f94f44243030ad51a10db6c9eb29 which relates to https://github.com/joyent/libuv/issues/526
22:33:26  <TooTallNate>tjfontaine: haha, looks like a red herring to me
22:33:33  <TooTallNate>did that make it into v0.8.8?
22:33:50  <tjfontaine>still checking that atm
22:34:35  <CIA-131>libuv: Ben Noordhuis v0.8 * r0ac2fdc / (include/uv.h src/unix/error.c): unix: map errno ESPIPE - http://git.io/KyOROw
22:34:44  <tjfontaine>TooTallNate: looks like it's only in libuv master atm
22:35:36  <bnoordhuis_>TooTallNate: what's the issue? you're not doing node < /dev/tty again, are you?
22:35:41  <tjfontaine>TooTallNate: for now you could just monkeypatch os.cpus to return something hacked
22:35:46  <tjfontaine>bnoordhuis_: worse, node on raspberry pi
22:37:57  <TooTallNate>the patch applies clean to v0.8.8 at least
22:39:28  <TooTallNate>bnoordhuis_: you're gonna love hearing about the openssl oddities i've been experiencing on this thing
22:39:33  <TooTallNate>and by love i mean… not
22:39:45  <bnoordhuis_>i don't doubt it
22:39:48  <bnoordhuis_>bundled openssl or?
22:39:59  <TooTallNate>both exhibit different weirdness
22:40:13  <bnoordhuis_>you're on your own with the system openssl
22:40:21  <bnoordhuis_>that's the generic you
22:40:27  * rendarquit
22:40:56  <TooTallNate>well with the bundled one, a hello world https server doesn't get a valid handshake
22:41:13  <tjfontaine>have fun debugging that
22:42:16  <TooTallNate>ya os.cpus() works now :D
22:42:28  <TooTallNate>well i know nothing about ssl internals, so any tips would be appreciated :)
22:42:43  <TooTallNate>when I link to the system openssl, it seems to work
22:43:01  <tjfontaine>TooTallNate: for simple starters test with openssl s_client -host <pi ip> -port <pi port>
22:43:05  <tjfontaine>and see what it says
22:44:46  <TooTallNate>this looks bad
22:44:47  <TooTallNate>1074558160:error:140790E5:SSL routines:SSL23_WRITE:ssl handshake failure:s23_lib.c:177:
22:45:29  <bnoordhuis_>TooTallNate: sanity check, what do you see when you `nc <ip> <port>`?
22:45:44  <TooTallNate>ohh now i got ben interested :)
22:45:49  * mikealquit (Quit: Leaving.)
22:46:20  <TooTallNate>bnoordhuis_: doesn't look like anything comes
22:46:29  <bnoordhuis_>what if you type something?
22:46:37  <tjfontaine>well, is it connected and can you do GET /\r\n\r\n
22:46:59  <bnoordhuis_>step 2 is to check with nmap -v -P0 -p<port> <ip>
22:48:06  <TooTallNate>tjfontaine: on port 80 i can, and that works
22:48:12  <TooTallNate>on 443, nothing
22:48:23  <TooTallNate>if i try to type something, the request closes
22:48:29  <TooTallNate>err, nc exits at least
22:48:32  <TooTallNate>nothing is printed
22:48:48  <tjfontaine>TooTallNate: buy me a pi and I'll help :)
22:49:08  <tjfontaine>I guess I could take some of the old routers and try them
22:49:23  <TooTallNate>i can foreword the port for a bit if you guys feel like digging in
22:51:07  <tjfontaine>I'm slightly curious
22:52:26  <TooTallNate>tjfontaine: you should be able to reach it at https://n8.io:4433
22:52:40  * bnoordhuis_is curious but busy
22:53:28  * TheJHquit (Ping timeout: 245 seconds)
22:53:46  <tjfontaine>hmm curious
22:53:53  <TooTallNate>bnoordhuis_: i ran nmap. seems like it ran without error. anything in particular I should be looking for?
22:55:34  * mikealjoined
22:55:52  <bnoordhuis_>TooTallNate: what did nmap report?
22:56:09  <TooTallNate>bnoordhuis_: https://gist.github.com/b7a1c1dac291b0acfbe5
22:56:45  <TooTallNate>is the odd port bypassing some SSL checks?
22:57:05  <tjfontaine>http://paste.debian.net/hidden/51efe08d/
23:00:07  <tjfontaine>TooTallNate: are you able to establish outbound https connections?
23:00:30  <TooTallNate>tjfontaine: trying
23:00:37  * travis-cijoined
23:00:37  <travis-ci>[travis-ci] joyent/libuv#614 (v0.8 - 0ac2fdc : Ben Noordhuis): The build was broken.
23:00:37  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/a7b83e0b9822...0ac2fdc55455
23:00:37  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2206562
23:00:37  * travis-cipart
23:02:23  <bnoordhuis_>TooTallNate: this is with the bundled openssl, right? for some reason your server doesn't speak tls, only ssl3
23:02:46  <tjfontaine>oh
23:03:18  <tjfontaine>how bout them apples, this isn't related to the cipher change is it?
23:03:30  <bnoordhuis_>unlikely
23:03:35  <bnoordhuis_>but easy to verify
23:03:52  <tjfontaine>indeed
23:04:34  <bnoordhuis_>TooTallNate: git show badbd1a | git apply -R; make
23:04:34  <tjfontaine>tls1 works, it's the ssl2 upgrade that seems to not
23:05:24  <tjfontaine>oh silly up not going back far enoguh
23:06:25  <tjfontaine>bnoordhuis_: hmm I'm not getting through on -ssl2 -ssl3 or -tls1
23:07:02  <bnoordhuis_>i can't test -ssl2, it doesn't work with my s_client
23:07:10  <bnoordhuis_>fun fact is that --help still prints the option :/
23:07:36  <tjfontaine>http://paste.debian.net/hidden/900bbd43/
23:07:55  <tjfontaine>end says "no cipher list" which is interesting
23:09:53  <bnoordhuis_>ah... maybe RC4 isn't compiled on arm?
23:09:58  <tjfontaine>quite possible
23:09:58  <bnoordhuis_>*compiled into openssl
23:10:18  * EhevuTovjoined
23:10:55  <bnoordhuis_>TooTallNate: what happens if you set the cipher list to e.g. AES128-SHA:AES256-SHA
23:14:05  * toothrchanged nick to toothrot
23:17:29  <isaacs>TooTallNate: you have n8.io? that's rad.
23:18:17  * piscisaureus_quit (Ping timeout: 265 seconds)
23:23:00  <TooTallNate>isaacs: ya it was a score :)
23:23:07  <TooTallNate>tjfontaine: bnoordhuis_: gonna try your suggestions now
23:26:45  * chobi_e_changed nick to chobi_e
23:29:35  <bnoordhuis_>ircretary: tell piscisaureus https://gist.github.com/79bded5c3bee9632d9b5 <- is there a reason to run close_socket() first? it's kind of the wrong thing to do on unices
23:29:35  <ircretary>bnoordhuis_: I'll be sure to tell piscisaureus
23:29:46  <bnoordhuis_>you're a doll, ircretary
23:30:26  <TooTallNate>tjfontaine: can't use tls outbound either https://gist.github.com/44bbc55f33b988e2a3c8
23:32:58  <ircretary>bnoordhuis: that's a little derrogatory. i'm a highly advanced robot, not a doll. i don't even have a body.
23:34:02  * chobi_echanged nick to chobi_e_
23:40:40  <bnoordhuis_>nice try, isaacs :)
23:41:41  <TooTallNate>bnoordhuis_: reverting badbd1a didn't seem to make a difference
23:43:20  * mmalecki_joined
23:45:32  * piscisaureus_joined
23:45:37  * mmaleckiquit (Ping timeout: 240 seconds)
23:46:15  <piscisaureus_>bnoordhuis_: no reason as far as i can remember
23:46:27  <piscisaureus_>bnoordhuis_: I think in uv.h I say that this is the wrong thing to do
23:46:44  <piscisaureus_>although, really, it doesn't matter here :)
23:47:16  <bnoordhuis_>piscisaureus_: it does in my epoll branch, it was failing with EBADF
23:47:27  <bnoordhuis_>i can handle that but it's really a bug in the test
23:47:35  <piscisaureus_>bnoordhuis_: agreed
23:47:55  <piscisaureus_>bnoordhuis_: although, I think libuv shouldn't crash if you can help it
23:48:31  <piscisaureus_>bnoordhuis_: if epoll returns ebadf, ideally, you'd figure out what the bad fd was and call the error cb
23:48:57  <bnoordhuis_>i can but EBADF nearly always means programmer error
23:49:30  * tomshredsjoined
23:49:39  <piscisaureus_>bnoordhuis_: True, I am in doubt. What does libev do? I think it handles ebadf
23:49:48  <bnoordhuis_>it does, for better or worse
23:49:51  <piscisaureus_>bnoordhuis_: at the very least, not a high priority issue :)
23:49:56  <bnoordhuis_>it was covering up a bug in uv-unix :)
23:50:08  <piscisaureus_>yeah, I've seen it happen on windows as well
23:50:23  <piscisaureus_>(the only bug that M.L. ever fixed for me)
23:50:30  <piscisaureus_>ok, going offline again
23:50:35  <bnoordhuis_>sleep tight, bertje
23:50:40  <bnoordhuis_>i may come over to 020 tomorrow
23:50:42  <piscisaureus_>my internet is down, everything is super slow
23:50:44  <piscisaureus_>oh
23:50:48  <piscisaureus_>bnoordhuis_: do it
23:50:52  <piscisaureus_>bnoordhuis_: noon-ish?
23:50:58  <bnoordhuis_>yes, if i feel like it :)
23:51:47  <piscisaureus_>bnoordhuis_: in my experience, if you start backing off at this point, you will definitely stay in 0182
23:51:55  <piscisaureus_>bnoordhuis_: anyway. see you
23:52:04  <bnoordhuis_>depends on how well i sleep
23:52:09  <bnoordhuis_>but sleep tight bertje
23:52:14  <piscisaureus_>thnx
23:54:14  <tjfontaine>TooTallNate: is this system or bundled openssl?
23:55:20  <TooTallNate>tjfontaine: bundled
23:56:20  <tjfontaine>quite interesting, I guess going to have to get one :)
23:57:01  * piscisaureus_quit (Ping timeout: 272 seconds)