00:02:55  <ryah>mjr_: make sure you have coredumps turned on on that efence node
00:03:08  <ryah>mjr_: it doesn't print any output - just segfaults early
00:03:22  <mjr_>we have cores enabled on all node processes now
00:03:28  <ryah>cool
00:03:36  <mjr_>So we'll catch any of them on any machine.
00:03:56  <ryah>i really can't wait until we can set the whole node-heap debugging problem onto dap's shoulders
00:04:44  <mjr_>men usage actually looks great now
00:10:57  <mjr_>ryah: that patch fails for me, is it against the v0.6 branch?
00:12:25  <ryah>mjr_: https://gist.github.com/e52c76f86e5cbd83afd4
00:12:28  <ryah>mjr_: yes
00:12:35  <ryah>i might have messed up the copy&paste
00:13:23  <pquerna>https://github.com/racker/node-failure-detector
00:14:26  * perezdquit (Quit: perezd)
00:14:30  <mjr_>ryah: I can dig into I guess. Just figured we must be looking at different trees
00:14:30  <mjr_>https://gist.github.com/339987ca182fac394c5f
00:14:55  <ryah>mjr_: that gist i just pasted is a new file - try that
00:15:10  <mjr_>That was the new one.
00:15:16  <mjr_>I'll figure it out.
00:15:21  <ryah>er
00:15:24  <ryah>do -p1
00:16:47  <CIA-111>node: Ryan Dahl master * rd85c85a / doc/api/addons.markdown : Change 'real example' in addon doc - http://git.io/I9j8GQ
00:16:47  <CIA-111>node: Ryan Dahl master * re4731f8 / (3 files in 3 dirs): Upgrade V8 to 3.8.2.1 - http://git.io/v4s1WQ
00:16:48  <CIA-111>node: Ryan Dahl master * rc2d9e62 / (11 files in 7 dirs):
00:16:48  <CIA-111>node: Merge remote branch 'origin/v0.6'
00:16:48  <CIA-111>node: Conflicts:
00:16:48  <CIA-111>node: deps/v8/src/version.cc - http://git.io/_X9kRA
00:16:52  <mjr_>yep, that was it. Ancient unix knowledge slowly fading from my brain.
00:17:02  * indexzerojoined
00:20:14  <mjr_>Do you expect that debugger repl test to fail?
00:20:33  <mjr_>https://gist.github.com/e838a8c38ef3a4258e4c
00:21:13  <ryah>dont worry about that - that test is flakey
00:21:18  <ryah>indutny--^
00:21:36  <ryah>indutny: especially on master - the debugger tests are leaking processes
00:24:38  * indexzeroquit (Quit: indexzero)
00:24:41  <mjr_>OK ryah, fire up your phone and watch as I restart your router process with this new patch
00:25:16  <ryah>i dont have my phone right now
00:25:37  <ryah>we decided to stop paying for internet and just use my cellphone tethering instead
00:25:41  <ryah>at home
00:25:53  <mjr_>oh man, how am I going to make you feel the pressure of a potentially broken patch then?
00:25:56  <ryah>which means.. i can't take my cellphone with me when lisa is at home
00:26:08  <mjr_>you guys share a phone?
00:26:18  <ryah>she has prepaid
00:26:22  <ryah>because she's poor
00:26:22  <ryah>:)
00:26:30  <mjr_>I'm going to buy you another phone for christmas
00:27:06  <ryah>we have many smartphones, just no service
00:27:16  * txdvquit (Remote host closed the connection)
00:27:23  <ryah>anyway - yes.
00:27:47  <ryah>brb
01:28:13  * sh1mmerquit (Quit: sh1mmer)
01:29:30  * travis-cijoined
01:29:31  <travis-ci>[travis-ci] joyent/node#142 (master - c2d9e62 : Ryan Dahl): The build is still failing.
01:29:31  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/624f70e...c2d9e62
01:29:31  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/442286
01:29:31  * travis-cipart
01:31:38  * perezdjoined
01:41:49  * pieternquit (Quit: pietern)
01:43:25  <CIA-111>node: Ryan Dahl master * r48a9a2d / lib/http.js : Add deprecation message for http.Client - http://git.io/xbC6xA
01:58:18  * travis-cijoined
01:58:19  <travis-ci>[travis-ci] joyent/node#143 (master - 48a9a2d : Ryan Dahl): The build is still failing.
01:58:19  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/c2d9e62...48a9a2d
01:58:19  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/442493
01:58:19  * travis-cipart
02:39:09  * TooTallNatequit (Quit: Linkinus - http://linkinus.com)
03:36:28  * mmaleckiquit (Ping timeout: 255 seconds)
04:26:55  * bbenviejoined
04:29:40  * benviequit (Ping timeout: 240 seconds)
04:44:55  <ryah>mjr_: how's that process?
05:06:09  <mjr_>lemme check. I think still good.
05:07:49  <mjr_>It hasn't crashed
05:08:02  <mjr_>We should roll this out to the cluster now and let it run for a day.
05:17:34  <ryah>ok
05:17:44  <ryah>crossing fingers
05:17:58  <mjr_>Any theories on the corruption thing?
05:18:14  <mjr_>I never did start one with efence. Getting too tired to attempt such a bold maneuver today.
05:26:17  <ryah>nope. heh - okay :)
05:27:09  <ryah>i agree, it is a bold maneuver
05:27:31  <ryah>i should benchmark that...
05:28:28  <indutny>ry§
05:28:34  <indutny>ryah: heya
05:28:40  <indutny>ryah: looking at the debugger
05:29:14  <indutny>ryah: btw have you thought about starting debuggee process as isolate?
05:29:20  <indutny>not as separate process
05:29:26  <ryah>indutny: yeah, we should do that once we land isolates
05:29:37  <indutny>ok, cool
05:29:56  <ryah>it would be nice if the V8 protocol could go over a pipe instead of TCP
05:30:02  <ryah>you should do a patch for that :)
05:30:35  <ryah>instead of specifying a port you should give it a socket fd.
05:30:37  <indutny>ryah: hehe :)
05:30:51  <ryah>on windows it can just use normal blocking winsock
05:30:52  <indutny>ryah: not sure that they'll apply it, but I'll try
05:31:15  <ryah>well - maybe you should allow V8 to create the pipe
05:31:21  <ryah>and return a handle to you
05:31:47  <indutny>ryah: hm... I've better idea
05:32:05  <indutny>ryah: just add write/read functions
05:32:13  <ryah>oh yeah - actually they have that
05:32:13  <indutny>so, yep like a pipe
05:32:14  <indutny>:D
05:32:20  <ryah>you dont need to do anything
05:32:49  <indutny>I think chrome isn't starting any debug processes
05:33:15  <ryah>just bind SendCommand() SetMessageHandler()
05:33:21  <ryah>^-- this is good
05:33:33  <indutny>yeah
05:34:37  <indutny>I think this is more priority than just fixing tests on master
05:34:40  <indutny>what is your opinion?
05:35:02  <indutny>ah, I can implement it in master first
05:35:22  <indutny>using fork
05:35:32  <ryah>yeah, do this on isolates2
05:35:36  <indutny>and sending/receving messages from child process
05:35:37  <ryah>it's going to get you familiar with our bugs
05:35:49  <indutny>ryah: ok, I'll see what I can do
05:36:06  <ryah>process.binding('isolates').create([args])
05:36:12  <ryah>that's all we have now
05:36:29  <ryah>adding a pipe for ipc now
05:36:54  <indutny>ryah: hm... why not use .fork()/
05:37:18  <indutny>ryah: and ask child isolate to call process._startDebugAgent()
05:37:42  <indutny>ryah: and then ask to invoke process._sendDebugMessage(...)
05:38:48  <ryah>sure
05:38:59  <indutny>ok
05:40:09  <indutny>I think I'll create separate node_debug.cc and node_debug.h for that purposes
06:55:39  <indutny>ryah: I'm creating node::Debug class
06:56:03  <indutny>ryah: and each isolate's process will get _debugger property which will be instance of node::Debug via ObjectWarp
06:56:27  <indutny>ryah: so Enable, Pause and anything will be methods of that class
06:56:30  <indutny>ryah: ok?
06:56:58  <indutny>ryah: then I'll just add interface for reading/writing to v8 debugger's queue
06:57:13  <indutny>ryah: and make i public for javascript side
06:57:22  <indutny>s/make i/make it/
07:07:07  <ryah>indutny: yeah - as long as it's unitialized when the debugger is off
07:07:22  <ryah>dont add a pointer in the C++
07:07:29  <ryah>use JS layer pointer
07:13:07  <indutny>ryah: huh?
07:13:16  <indutny>ryah: what do you mean by saying pointer?
07:14:14  <indutny>ah, you want it to not affect startup time?
07:14:48  <indutny>ok, process._debugger = { init: function() { process._debugger = new binding.Debug })
07:14:56  <indutny>something like that? ;)
07:25:16  * Raynosquit (Remote host closed the connection)
07:54:40  * Raynosjoined
07:56:14  <indutny>ryah: you can follow up progress here: https://github.com/indutny/node/tree/feature-isolate-debugger, if you're interested
08:40:23  * mmaleckijoined
11:35:39  * mralephjoined
11:39:13  * txdvjoined
12:12:30  * paddybyersquit (Quit: paddybyers)
12:14:21  * paddybyersjoined
12:14:43  * paddybyerspart
13:14:46  * piscisaureus_joined
13:17:48  * piscisaureus__joined
13:20:35  * piscisaureus_quit (Ping timeout: 244 seconds)
13:34:04  * mmaleckichanged nick to mmalecki[xmas]
16:13:33  <indutny>ryah: hey man
16:13:34  <indutny>yt?
16:43:49  <indutny>nvm
16:45:42  * paddybyersjoined
16:48:36  * piscisaureus_joined
17:09:38  <txdv>marry chgristmas
17:10:00  <txdv>and thanks for thge awesome library you have distilled out of nodejs so others could enjoy programming too :)
17:34:16  * piscisaureus_quit (Ping timeout: 252 seconds)
17:35:00  * paddybyersquit (Quit: paddybyers)
18:26:01  * igorzijoined
18:27:36  * dshaw_joined
18:39:35  * sh1mmerjoined
19:21:41  * mmalecki[xmas]changed nick to mmalecki
19:39:15  * piscisaureus_joined
20:07:55  * perezdquit (Quit: perezd)
20:09:33  * dshaw_quit (Quit: Leaving.)
20:18:49  * piscisaureus_quit (Ping timeout: 248 seconds)
20:25:23  * igorziquit (Ping timeout: 258 seconds)
20:38:45  <indutny>marry xmas everyone!
20:39:09  <indutny>ryah: SetDebugMessageDispatchHandler <- why are we using this?
20:39:25  <indutny>was that made because of some v8 bug?
20:39:41  <indutny>it's causing a lot of problems with isolates
20:55:54  * paddybyersjoined
21:00:39  * ericktjoined
21:19:40  * ericktquit (Quit: erickt)
22:14:25  <ryah>indutny: i dont remember
22:14:53  <ryah>mjr_: any crashes?
22:19:28  * AndreasMadsenjoined
22:26:06  * AndreasMadsenquit (Quit: Leaving)
22:26:16  * AndreasMadsenjoined
22:34:12  <mjr_>ryah: none in that process. I need to push it out to more machines, but I haven't yet.
22:36:57  <mjr_>I guess 8 processes have been running that patch for a while, and one of them did core.
22:36:58  <mjr_>https://gist.github.com/a441348af7f643676969
22:37:45  <mjr_>So this is pretty clearly our audio data getting overwritten into some http headers, but I still can't tell if node is doing this or if we are sending it up this way.
22:55:08  * AndreasM_joined
22:55:26  * AndreasMadsenquit (Ping timeout: 252 seconds)
23:02:08  * paddybyersquit (Quit: paddybyers)
23:05:12  * AndreasM_quit (Ping timeout: 268 seconds)
23:28:45  * piscisaureus_joined
23:31:47  * AndreasMadsenjoined
23:33:36  <ryah>mjr_:please do the efence thing
23:35:52  * ericktjoined
23:37:35  * ericktquit (Client Quit)
23:37:46  <mmalecki>ryah: hey, so I've read a bit about linking stuff. I can actually understand your concerns now.
23:38:04  <mmalecki>it should be easier.
23:41:11  <ryah>mmalecki: good :)
23:41:28  <ryah>piscisaureus_: how's the linked-list coming?
23:47:07  * AndreasMadsenquit (Ping timeout: 252 seconds)
23:47:08  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
23:51:17  * AndreasMadsenjoined
23:59:02  * bbenviequit