00:04:03  <mmalecki>bnoordhuis: would it be a good idea to define a UV_PATH_MAX which would contain maximum file name length for each OS?
00:04:25  <bnoordhuis>mmalecki: no
00:04:35  <bnoordhuis>for a number of reasons
00:04:47  <mmalecki>bnoordhuis: care to explain?
00:04:52  <bnoordhuis>sure
00:05:00  <bnoordhuis>maximum file name length for each OS <- there is no such thing, it's fs dependent
00:05:28  <bnoordhuis>it's even conceivable that a file system doesn't have path limits
00:05:35  <mmalecki>bnoordhuis: I mean, Unix has MAX_PATH defined
00:05:43  <bnoordhuis>yes, a mistake from the past
00:05:51  <mmalecki>oh well, ok
00:06:04  <mmalecki>so how do I write a test case for ENAMETOOLONG?
00:06:27  <bnoordhuis>well, it's complicated
00:06:37  <mmalecki>making a 1024 char string is enough?
00:06:39  <bnoordhuis>most file systems have a 1k or 4k limit
00:06:57  <mmalecki>ok, I'll just make 8k
00:07:10  <raggi>there's functions in posix and bsd for this, not sure about linux
00:07:25  <raggi>pathconf, sysconf, etc expose stuff
00:07:47  <mmalecki>raggi: oh, I see
00:07:57  <mmalecki>bnoordhuis: can we use ^ ?
00:08:10  <raggi>n.b. the functions may lie
00:08:15  <mmalecki>`man pathconf` <- in case you're too lazy
00:08:30  <bnoordhuis>mmalecki: what raggi said
00:08:44  <mmalecki>ok, let's spam it with 8k chars.
00:09:08  <mmalecki>it's just a case, no need to make it work on every single system out there
00:09:14  * paddybyersquit (Quit: paddybyers)
00:09:55  <raggi>mmalecki: pathconf takes a path, so you "should" use it each time you want to write into a dir
00:10:20  <bnoordhuis>mmalecki: fun fact about pathconf: it can return -1 to indicate that it's unbounded
00:11:07  <bnoordhuis>there is a comment somewhere in libuv where i wax philosophically about that
00:11:40  <mmalecki>bnoordhuis: so fun
00:12:07  <raggi>https://gist.github.com/00db15a1411bb1230b96
00:12:08  <raggi>f.e.
00:36:04  <mmalecki>yay, looks like I got the ENAMETOOLONG mapping ok
00:36:17  <mmalecki>bnoordhuis: hey, what was the command to update libuv in node, again?
00:36:29  <mmalecki>bnoordhuis: actually, nvm, I'll just copy it XD
00:37:56  <mmalecki>bnoordhuis: against which branch should I pull request?
00:42:18  <bnoordhuis>mmalecki: if you're patching both node and libuv, you'll have to split it into two PRs
00:42:28  <bnoordhuis>what are you hacking on?
00:42:56  <mmalecki>bnoordhuis: not patching both of them. just making ENAMETOOLONG mapping
00:43:26  <mmalecki>bnoordhuis: that goes to libuv, of course
00:43:37  <mmalecki>bnoordhuis: I just wanted to test it with node
00:44:08  <bnoordhuis>mmalecki: btw, `rm -rf deps/uv/ && (cd ../libuv && git archive --format=tar --prefix=deps/uv/ HEAD) | tar x`
00:44:49  <mmalecki>bnoordhuis: thanks
00:47:15  <mmalecki>bnoordhuis: so, which branch?
00:47:55  * mikealquit (Quit: Leaving.)
00:49:18  <bnoordhuis>mmalecki: is it a bug fix or a new feature?
00:49:50  <mmalecki>bnoordhuis: bug fix: https://github.com/joyent/libuv/issues/247
00:49:54  <mmalecki>bnoordhuis: so, v0.6?
00:50:18  <bnoordhuis>mmalecki: yes
00:52:05  <mmalecki>bnoordhuis: hm. one weird thing
00:52:19  <mmalecki>bnoordhuis: looks like in windows, this error is called ERROR_BUFFER_OVERFLOW
00:52:44  <mmalecki>bnoordhuis: is it reasonable to map it to UV_ENAMETOOLONG as well? this name seems vague.
00:53:09  <bnoordhuis>mmalecki: you should ask bert or igor
00:53:30  <bnoordhuis>though i think ERROR_BUFFER_OVERFLOW is a generic error code
00:53:39  <mmalecki>bnoordhuis: http://msdn.microsoft.com/en-us/library/windows/desktop/ms681382(v=vs.85).aspx
00:53:45  <mmalecki>Ctrl+F ERROR_BUFFER_OVERFLOW
00:54:16  <mmalecki>bnoordhuis: ok, I have nothing to test this thing on anyway
00:54:25  <mmalecki>I mean, no windows machine
00:54:36  <mmalecki>I'll just pull request unix stuff
00:58:17  * mikealjoined
00:59:06  <bnoordhuis>sleep tight, guys
00:59:23  <mmalecki>bnoordhuis: night!
00:59:29  * bnoordhuisquit (Quit: Leaving)
01:13:16  * mikealquit (Quit: Leaving.)
01:25:37  * piscisaureus_joined
01:29:47  * AvianFluquit (Ping timeout: 255 seconds)
01:33:46  * mikealjoined
01:35:35  <benvie>this js might be some limited use, a js mapping of the windows error codes https://github.com/Benvie/node-ffi-tools/blob/master/win32/errors.js
01:38:44  <mmalecki>benvie: oh, thanks
01:38:59  <mmalecki>these all errors are so vague to me
01:39:13  <benvie>yeah, a handful have comments but few do
01:40:05  <benvie>I plan to go through and attach one liner comments to the ones that seem to happen commonly, drawing from the MS docs
01:42:47  * AvianFlujoined
01:47:20  * mikealquit (Quit: Leaving.)
01:48:03  <mmalecki>benvie: that'd be nice
01:48:16  <mmalecki>anyways, pull request done, time to go to sleep
01:48:19  <benvie>night
01:48:34  <mmalecki>night!
01:50:50  * mmaleckichanged nick to mmalecki[zzz]
01:57:50  * piscisaureus_quit (Read error: Connection reset by peer)
02:07:12  * creationixpart
02:19:09  * mikealjoined
02:21:03  * mikealquit (Client Quit)
02:21:45  * mikealjoined
02:24:02  * brsonquit (Quit: leaving)
02:58:30  * mikealquit (Quit: Leaving.)
03:06:24  * isaacsquit (Remote host closed the connection)
03:14:51  * mikealjoined
03:56:47  * dshaw_joined
03:58:06  * mikealquit (Quit: Leaving.)
04:07:32  * bmeck_vacationquit (Ping timeout: 276 seconds)
04:14:22  * isaacsjoined
04:32:42  * bmeck_vacationjoined
04:40:45  * orlandovftwjoined
04:55:31  * perezdjoined
04:55:48  * bmeck_vacationquit (Ping timeout: 260 seconds)
05:10:00  * sh1mmerquit (Quit: sh1mmer)
05:10:30  * sh1mmerjoined
05:28:40  * orlandovftwquit (Ping timeout: 240 seconds)
05:29:55  * perezdquit (Quit: perezd)
05:57:23  * isaacsquit (Remote host closed the connection)
06:00:47  <CIA-115>node: koichik v0.6 * r3df7c90 / (lib/http.js test/simple/test-http-should-keep-alive.js): (log message trimmed)
06:00:47  <CIA-115>node: http: keep-alive should default with HTTP/1.1 server
06:00:47  <CIA-115>node: As RFC 2616 says we should, assume that servers will provide a persistent
06:00:47  <CIA-115>node: connection by default.
06:00:47  <CIA-115>node: > A significant difference between HTTP/1.1 and earlier versions of
06:00:47  <CIA-115>node: > HTTP is that persistent connections are the default behavior of any
06:00:48  <CIA-115>node: > HTTP connection. That is, unless otherwise indicated, the client
06:03:01  * isaacsjoined
06:08:34  * travis-cijoined
06:08:34  <travis-ci>[travis-ci] joyent/node#296 (v0.6 - 3df7c90 : koichik): The build was broken.
06:08:34  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/f2b1f57...3df7c90
06:08:34  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/556229
06:08:34  * travis-cipart
06:09:03  * mikealjoined
06:10:44  * isaacsquit (Remote host closed the connection)
06:42:53  * isaacsjoined
07:05:09  * isaacsquit (Remote host closed the connection)
07:05:41  * isaacsjoined
07:09:59  * isaacsquit (Ping timeout: 255 seconds)
07:10:26  * mikealquit (Quit: Leaving.)
07:11:27  * AvianFluquit (Read error: Connection reset by peer)
07:13:02  * mikealjoined
07:19:17  * mikealquit (Quit: Leaving.)
07:29:55  * mikealjoined
07:38:49  * orlandovftwjoined
07:48:19  * AvianFlujoined
07:53:33  * indutny_sleepingchanged nick to indutny
08:36:13  * sh1mmerquit (Quit: sh1mmer)
08:41:53  <CIA-115>node: Brandon Benvie v0.6 * rf901443 / lib/util.js : (log message trimmed)
08:41:53  <CIA-115>node: util: use getOwnPropertyDescripter
08:41:53  <CIA-115>node: Change formatProperty in util.js to use Object.getOwnPropertyDescriptor
08:41:53  <CIA-115>node: instead of __lookup[GS]etter__.
08:41:53  <CIA-115>node: Use the cached value from the descriptor to reduce number of property
08:41:53  <CIA-115>node: lookups from 3 to 1.
08:41:54  <CIA-115>node: Fallback to standard lookup if the descriptor is empty. This doesn't
08:41:54  <CIA-115>node: koichik v0.6 * r93298af / test/simple/test-util-inspect.js : test: for #2109 - http://git.io/S-e8Wg
08:49:38  * travis-cijoined
08:49:38  <travis-ci>[travis-ci] joyent/node#297 (v0.6 - 93298af : koichik): The build was fixed.
08:49:38  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/3df7c90...93298af
08:49:38  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/556525
08:49:38  * travis-cipart
08:53:35  * mikealquit (Quit: Leaving.)
08:55:42  * mikealjoined
09:03:23  * orlandovftwquit (Ping timeout: 255 seconds)
09:04:51  * mralephjoined
09:33:32  * paddybyersjoined
09:34:20  * paddybyersquit (Client Quit)
11:33:07  * AndreasMadsenjoined
11:33:35  * AndreasMadsenquit (Remote host closed the connection)
11:34:24  * AndreasMadsenjoined
11:38:48  <AndreasMadsen>indutny: Okay the process._channel.close() is definitely sync.
11:38:48  <AndreasMadsen>I will update #2426 to match the current master, when that has landed I will look intro the cluster problem.
11:38:55  * AndreasMadsenpart
11:39:19  * AndreasMadsenjoined
11:39:23  * AndreasMadsenquit (Remote host closed the connection)
12:10:14  <indutny>AvianFlu: cool!
12:10:20  <indutny>oop
12:10:26  <indutny>wrong nick
12:10:27  <indutny>sorry
12:10:30  <indutny>AndreasMadsen ^
12:37:32  <mmalecki[zzz]>morning!
12:37:33  * txdvquit (Read error: Connection reset by peer)
12:38:06  * txdvjoined
12:38:40  * mmalecki[zzz]changed nick to mmalecki
13:16:25  * paddybyersjoined
14:07:18  * paddybyersquit (Quit: paddybyers)
14:34:31  <CIA-115>node: koichik v0.6 * r8271800 / test/simple/test-net-server-max-connections.js :
14:34:31  <CIA-115>node: test: fix test/simple/test-net-server-max-connections.js is racey
14:34:31  <CIA-115>node: Fixes #1333. - http://git.io/B1DMZg
14:42:12  * travis-cijoined
14:42:12  <travis-ci>[travis-ci] joyent/node#298 (v0.6 - 8271800 : koichik): The build was broken.
14:42:12  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/93298af...8271800
14:42:12  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/557609
14:42:12  * travis-cipart
15:48:27  * AndreasMadsenjoined
16:34:43  * dshaw_quit (Quit: Leaving.)
16:36:53  <AndreasMadsen>Is there a 0.6 -> 0.7 migration document
16:41:40  <mmalecki>AndreasMadsen: yes, in the Wiki
16:41:42  <mmalecki>I started it
16:43:08  <AndreasMadsen>mmalecki: oh, nice I will properly add some cluster stuff later :) - or :( because it is necessary
16:43:26  <mmalecki>AndreasMadsen: API changes are good :)
16:44:33  <AndreasMadsen>mmalecki: required API additions are, I'm not much for the changes. But as they say - live on the edge
16:49:51  <AndreasMadsen>Oh forgot to say:
16:49:52  <AndreasMadsen>yo
16:50:22  <mmalecki>ohai
16:55:47  * bnoordhuisjoined
17:00:21  <mmalecki>bnoordhuis: hey
17:00:34  <bnoordhuis>mmalecki: ho
17:00:48  <mmalecki>bnoordhuis: my first real libuv patch :3 https://github.com/joyent/libuv/pull/295
17:01:00  <mmalecki>bnoordhuis: mind reviewing?
17:01:11  <mmalecki>bnoordhuis: like, when you have time to do it?
17:01:14  <bnoordhuis>mmalecki: i've seen it
17:01:27  <bnoordhuis>still recuperating
17:01:39  <mmalecki>lol
17:02:07  <mmalecki>(really?)
17:02:12  <bnoordhuis>i kid, it looks fine to me
17:02:17  <mmalecki>yay
17:02:49  <bnoordhuis>i'm going to let bert chime in on monday, he probably knows what the proper approach is on windows
17:03:13  <mmalecki>bnoordhuis: can you merge it for Unix?
17:03:36  <bnoordhuis>are you in a rush?
17:03:40  <mmalecki>bnoordhuis: or do you have to wait for windows? not sure what the policy is?
17:03:47  <mmalecki>bnoordhuis: nah, just curious
17:03:47  <bnoordhuis>oh, like that
17:04:05  <bnoordhuis>i preferably only merge patches that work on all platforms
17:04:35  <bnoordhuis>if i merge it now and forget about it, someone is going to post a windows issue in two weeks time
17:04:59  <mmalecki>bnoordhuis: I see, thanks! :)
17:06:49  <mmalecki>bnoordhuis: I wonder, is the a shorter way to do https://github.com/joyent/libuv/pull/295/files#L2R459 ?
17:08:31  <mmalecki>bnoordhuis: also, does https://github.com/mmalecki/node/commit/030d20c47f7ad2997a354a1bf060d07267df3e7e follow node c++ style?
17:08:32  <bnoordhuis>mmalecki: populating the array, you mean? not really
17:08:33  <AndreasMadsen>mmalecki: bnoordhuis: is the UV_ERRNO_MAP exposed to node.js in some binding? It would be very usefull
17:09:13  <mmalecki>AndreasMadsen: I don't think so... but you're right
17:09:45  <bnoordhuis>mmalecki: i try to keep the number of #ifdefs to a minimum, they make code hard to reason about
17:10:20  <AndreasMadsen>mmalecki: yes, The errnoException could then just use that map to make errno code readable
17:11:00  <bnoordhuis>mmalecki: in that particular case, i'd probably just remove the line altogether
17:11:07  <bnoordhuis>retrying is not really feasible at that point
17:11:17  <mmalecki>bnoordhuis: it's needed later, when printing debug strings
17:11:32  <bnoordhuis>AndreasMadsen: UV_ERRNO_MAP - no
17:11:52  <AndreasMadsen>bnoordhuis: that made me sad :(
17:11:52  <mmalecki>https://github.com/mmalecki/node/blob/030d20c47f7ad2997a354a1bf060d07267df3e7e/src/node_crypto.cc#L614 and https://github.com/mmalecki/node/blob/030d20c47f7ad2997a354a1bf060d07267df3e7e/src/node_crypto.cc#L618
17:12:55  <bnoordhuis>AndreasMadsen: but probably not as sad as run over kittens
17:13:16  <bnoordhuis>mmalecki: are we talking about the same thing?
17:13:27  <mmalecki>or these kids bnoordhuis used to beat up everyday on his way to work
17:14:09  <mmalecki>bnoordhuis: not sure, I'm talking about removing unused variable warning, when SSL_PRINT_DEBUG is not defined
17:14:35  <AndreasMadsen>bnoordhuis: It is only when I see nodejitsu do process.exit(0) after child_process.fork(), I see kittens die
17:14:55  <mmalecki>:)
17:15:18  <indutny>bnoordhuis: heya
17:15:29  <mmalecki>AndreasMadsen: I think we can just fork._channel.close() now, right?
17:15:54  <mmalecki>(I *really* want to get rid of daemon.node)
17:16:05  <indutny>bnoordhuis: vflash is kinda ignoring us :D
17:16:11  <bnoordhuis>mmalecki: retry is not just used, the BIO_should_retry() call is superfluous
17:16:13  <indutny>bnoordhuis: but I think I answered his questions
17:16:18  <indutny>bnoordhuis: if you want, I can translate them
17:16:31  <bnoordhuis>indutny: what was it about?
17:17:01  <mmalecki>bnoordhuis: it is, when printing some debug strings there. (when SSL_PRINT_DEBUG is defined)
17:17:04  <AndreasMadsen>mmalecki: no, you need much more. You should implement the isolate function from: https://github.com/joyent/node/pull/2426/files
17:17:33  <mmalecki>AndreasMadsen: ok, I guess daemon.node is not that bad :)
17:17:42  <bnoordhuis>mmalecki: oh, i see what you mean
17:17:48  <indutny>bnoordhuis: about zlib leak, he said that example in issue doesn't work for him on 0.6.x
17:17:50  <mmalecki>AndreasMadsen: looking forward to your .disconnect patch, tho
17:17:51  <bnoordhuis>mmalecki: easy fix: (void) retry;
17:17:52  <indutny>bnoordhuis: and on 0.7.x too
17:17:56  * isaacsjoined
17:18:15  <AndreasMadsen>mmalecki: I'm updating it right now :)
17:18:31  <mmalecki>AndreasMadsen: awesome! :). link to pull request?
17:18:48  <AndreasMadsen>mmalecki: old one -> https://github.com/joyent/node/pull/2426
17:19:21  <AndreasMadsen>But I will properly spilt it up in two pull requester later today :)
17:19:28  <mmalecki>AndreasMadsen: k, sounds good
17:20:29  <CIA-115>node: Ben Noordhuis v0.6 * r9edb984 / src/node_crypto.cc :
17:20:30  <CIA-115>node: crypto: silence unused variable warning
17:20:30  <CIA-115>node: `retry` is not used if SSL_PRINT_DEBUG is not defined. - http://git.io/QZbNfg
17:20:33  <bnoordhuis>mmalecki: ^
17:20:55  <mmalecki>bnoordhuis: fun trick you have there
17:20:58  <bnoordhuis>indutny: oh okay
17:22:23  <AndreasMadsen>could the core team merge v0.6 intro master - I need some updates from it :)
17:23:03  <indutny>bnoordhuis: could I ?:)
17:23:12  <indutny>I kinda like merging
17:23:19  <indutny>(well, sometimes)
17:23:23  <bnoordhuis>indutny: go ahead
17:23:41  <indutny>k
17:23:57  <mmalecki>oh fuck you waf
17:24:38  <mmalecki>it breaks all unix conventions :/
17:25:04  <mmalecki>make -j8 won't work because waf is retarded :/
17:26:34  <bnoordhuis>mmalecki: `JOBS=8 make`
17:26:57  <mmalecki>bnoordhuis: I'm aware. it gets annoying if you have to do it in a script
17:27:13  <mmalecki>ah, there's no wscript on master. I can detect it, cool
17:28:19  * travis-cijoined
17:28:19  <travis-ci>[travis-ci] joyent/node#299 (v0.6 - 9edb984 : Ben Noordhuis): The build was fixed.
17:28:19  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/8271800...9edb984
17:28:19  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/558159
17:28:19  * travis-cipart
17:31:10  <AndreasMadsen>indutny: the child._channel.close(); is sync, the problem is just that there is no async way to detect that the channel is closed in the child. - Or parent if process._channel.close() was called.
17:31:53  <indutny>AndreasMadsen: ok
17:32:07  <indutny>AndreasMadsen: looks like cluster has conflicting changes in v0.6 and master
17:33:15  <AndreasMadsen>indutny: it is properly a fix in uncaughtException - it is fixed in the master, but I do not know :/
17:33:28  <indutny>AndreasMadsen: ok
17:36:29  <indutny>AndreasMadsen: test/simple/test-cluster-kill-workers.js is deleted in master
17:36:36  <indutny>AndreasMadsen: ok?
17:37:48  <AndreasMadsen>indutny: sec
17:38:56  <AndreasMadsen>indutny: There is another testcase for that, I remember that I would it ustable - so that is ok.
17:40:46  <indutny>bnoordhuis: yt?
17:41:20  <indutny>bnoordhuis: I first time used merge -s ours
17:41:40  <indutny>bnoordhuis: so I'm a little bit worried about results, because there're no conflicts at all
17:45:07  <indutny>bnoordhuis: https://github.com/indutny/node/compare/master...merge-0.6-1
17:46:44  <indutny>ok, looks like I merged changes from libuv too
17:46:46  <indutny>my ba
17:46:47  <AndreasMadsen>oh, forget about "the child._channel.close(); is sync, the problem is just that there is no async way to detect that" just solved it :)
17:46:50  <indutny>s/ba/bad
17:46:57  <indutny>AndreasMadsen: cool
17:59:30  <indutny>bnoordhuis: will udp multicast features work w/o uv update?
18:02:37  <mmalecki>can anyone repro race condition in child-process-fork2?
18:02:46  <mmalecki>working on fixing it
18:02:57  * mikealquit (Quit: Leaving.)
18:03:17  <indutny>looks like yes
18:03:22  <indutny>bnoordhuis: ^
18:03:55  <indutny>bnoordhuis: can you merge uv first?
18:05:05  <AndreasMadsen>mmalecki: acturrly I have never encountered it my self, and I have executed child-process testcases too many times
18:05:22  <AndreasMadsen>s/acturrly/actually
18:06:04  * mikealjoined
18:08:39  <mmalecki>I can repro with lot of CPU usage
18:08:49  <mmalecki>as in "makes MBP cry"
18:33:11  * mikealquit (Quit: Leaving.)
18:50:43  <AndreasMadsen>mmalecki: indutny: disconnect method patch: https://github.com/joyent/node/pull/2591
18:53:07  <mmalecki>AndreasMadsen: neat! except grammar in docs ;)
18:59:35  <AndreasMadsen>mmalecki: oh, this danish language that only 5.5 million know it totally useless :)
19:00:03  <indutny>AndreasMadsen: learn russian (150m) or chineese :D
19:01:16  <mmalecki>polish is 40m, I think
19:01:59  <mmalecki>but yeah, well, I find it useless these days :)
19:03:42  <indutny>AndreasMadsen: so basically, it'll notify child that'll die soon and close IPC channel (so further messages won't pass)
19:04:02  <indutny>AndreasMadsen: right?
19:04:40  <AndreasMadsen>indutny: it will just close the IPC channel, nobody says that it will die soon, except process.on('exit')
19:04:45  <indutny>AndreasMadsen: what if child was in middle of writing something to IPC
19:05:06  <indutny>AndreasMadsen: probably better stop accepting "new" messages?
19:05:23  <mmalecki>indutny: process.std{out,in} are synchronous
19:05:26  <indutny>or close once message received
19:05:33  <indutny>mmalecki: and what?
19:05:45  <indutny>mmalecki: receving data is still asynchronous
19:05:50  <AndreasMadsen>mmalecki: not between parent and child
19:06:13  <indutny>AndreasMadsen: so for example child is sending 10gb object to parent
19:06:14  <mmalecki>AndreasMadsen: oh, nice, didn't know that
19:06:28  <indutny>AndreasMadsen: and in the middle of this parent is calling .dsconnect()
19:06:44  <indutny>AndreasMadsen: parent should receive thismessage first
19:06:50  <indutny>and only then close pipe
19:08:25  <indutny>AndreasMadsen: is that clear?
19:08:29  <AndreasMadsen>indutny: I see you point, but how should this be accompliced
19:08:57  <indutny>AndreasMadsen: you should be aware of receive state
19:09:24  <indutny>AndreasMadsen: if some data is already queued in buffer it'll invoke https://github.com/AndreasMadsen/node-1/blob/95f0d021aee9dcdce00f996dafb7dc37a3f9fa54/lib/child_process.js#L88 anyway
19:09:31  * piscisaureus_joined
19:10:16  <indutny>AndreasMadsen: but if you're in the middle of thing - wait for '\n', emit 'message'
19:10:22  <indutny>AndreasMadsen: and close channel
19:11:09  <indutny>AndreasMadsen: but you should send 'disconnect' to child immediately anyway
19:11:23  <indutny>piscisaureus_: hi
19:11:35  <piscisaureus_>indutny: sup?
19:11:36  <indutny>mmalecki: MVCC in bplus :D
19:11:41  <indutny>piscisaureus_: just hi
19:11:41  <indutny>:D
19:11:51  <mmalecki>indutny: w00t w000t
19:11:52  <piscisaureus_>indutny: hi to you :-)
19:12:40  <AndreasMadsen>indutny: so we should send an internal IPC message to (the other side) where the disconnected flag will be true and 'disconnect' is emitted, when the readQuery is empty we will close the channel.
19:12:45  <AndreasMadsen>Right?
19:14:08  <AndreasMadsen>And then the disconnect event will be emitted in parrent.
19:14:13  * piscisaureus_afm
19:14:15  <piscisaureus_>afk
19:14:25  <AndreasMadsen>s/parrent/parent/
19:15:17  <indutny>AndreasMadsen: messaging is ok in your patch, do not add anything new. just wait for newline symbol if `jsonBuffer` is not empty and only after '\n' close channel
19:15:18  <AndreasMadsen>But what if the parent is sending a message between the .disconnect call and the disconnect emit?
19:15:42  <indutny>AndreasMadsen: it should not be able to do this
19:16:01  <indutny>AndreasMadsen: I think you can filter this here: https://github.com/AndreasMadsen/node-1/blob/95f0d021aee9dcdce00f996dafb7dc37a3f9fa54/lib/child_process.js#L126
19:17:41  <AndreasMadsen>indutny: but there could be more that one message in the query poll
19:18:11  <AndreasMadsen>Will we then just ignore the rest?
19:18:29  <indutny>AndreasMadsen: I think yes
19:18:38  <indutny>AndreasMadsen: I don't want to add a lot of complexity
19:18:46  <indutny>AndreasMadsen: but if it's simple to check - just do it ;)
19:18:53  <indutny>gtg
19:18:54  <indutny>ttyl
19:19:20  <AndreasMadsen>indutny: i need to look intro it :)
19:22:34  <mmalecki>hrm...
19:22:43  <mmalecki>we've had 0.7.0 release, right?
19:23:19  <mmalecki>https://github.com/joyent/node/blob/master/src/node_version.h#L31
19:23:25  <mmalecki>I think it should be 1 now?
19:24:42  * dshaw_joined
19:26:18  <mmalecki>indutny: also, ping :)
19:31:22  <piscisaureus_>it's weekend guys. You should spend time with your family or sit in a bar :-)
19:31:50  <mmalecki>piscisaureus_: drinking home today :/
19:34:35  <piscisaureus_>so sad
19:35:17  <mmalecki>yeah. so sad I think about actually starting
19:35:32  * benviequit
19:38:05  <mmalecki>BUT REALLY VERSION NUMBER IS WRONG :(
19:38:21  <mmalecki>it's just plain wrong
19:43:20  <isaacs>mmalecki: yeah, looks that way
19:43:59  <mmalecki>isaacs: thank you. I'll be able to fall asleep.
19:44:51  <isaacs>mmalecki: this is missing the 0->1 update: https://github.com/joyent/node/commit/72cbb7492cb5956c8265b415bd6ff473b7b89f35
19:44:57  <isaacs>mmalecki: pr?
19:45:00  * piscisaureus_quit (Ping timeout: 240 seconds)
19:45:02  <mmalecki>isaacs: yeah, I can do that
19:45:09  <isaacs>mmalecki: you don't have to.
19:45:13  <isaacs>if you'd rather drink or sleep or something :)
19:45:24  <mmalecki>isaacs: I'll. I want to acquire core commits
19:45:28  <mmalecki>(fighting with indutny)
19:49:33  * TooTallNatejoined
19:54:18  * benviejoined
19:54:45  <isaacs>hahah, good luck on that one
19:54:47  <mmalecki>isaacs: https://github.com/joyent/node/pull/2592
19:54:56  <isaacs>sorry, i can't reveiw patches today
19:55:00  <isaacs>it's the sabbath.
19:55:13  <mmalecki>isaacs: sure, np
19:55:21  <mmalecki>BUT I'LL BEAT HIM!
19:55:23  <mmalecki>one day.
19:56:35  <isaacs>(few atheists take advantage of the fact that we can basically make up our own religious rules any time we want.)
19:57:02  <mmalecki>lol
19:58:57  <mmalecki>hm, mutable religion. nice.
19:59:44  * piscisaureus_joined
20:05:35  * TooTallNatequit (Quit: Linkinus - http://linkinus.com)
20:16:04  * mralephquit (Quit: Leaving.)
20:22:49  * AndreasMadsenquit (Remote host closed the connection)
20:51:53  * mralephjoined
20:57:49  * mralephquit (Quit: Leaving.)
21:42:43  * brsonjoined
21:56:07  * piscisaureus_quit (Read error: Connection reset by peer)
21:56:47  * piscisaureus_joined
22:02:49  * piscisaureus_quit (Ping timeout: 272 seconds)
22:09:26  * piscisaureus_joined
22:13:53  <CIA-115>node: Maciej Małecki master * rff51263 / src/node_version.h :
22:13:53  <CIA-115>node: version: bump `NODE_PATCH_VERSION`
22:13:53  <CIA-115>node: Patch version hasn't been bumped correctly in
22:13:53  <CIA-115>node: 72cbb7492cb5956c8265b415bd6ff473b7b89f35. - http://git.io/vozfuw
22:27:35  * travis-cijoined
22:27:35  <travis-ci>[travis-ci] joyent/node#300 (master - ff51263 : Maciej Małecki): The build is still failing.
22:27:35  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/6d00c08...ff51263
22:27:35  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/559427
22:27:35  * travis-cipart
23:08:02  * TooTallNatejoined
23:18:02  * piscisaureus_quit (Ping timeout: 252 seconds)
23:18:41  * piscisaureus_joined
23:18:56  * mikealjoined
23:34:31  * TooTallNatequit (Quit: Linkinus - http://linkinus.com)
23:36:13  * txdvchanged nick to a
23:36:31  * achanged nick to txdv
23:37:14  * txdvchanged nick to txdva
23:38:19  * txdvachanged nick to txdv
23:38:34  * txdvchanged nick to txdv2
23:38:50  * txdv2changed nick to txdv
23:39:29  * txdvchanged nick to txdv2
23:42:28  * mikealquit (Quit: Leaving.)
23:45:39  * mikealjoined
23:54:51  * piscisaureus__joined
23:57:05  * `3rdEdenjoined