00:03:46  * dapquit (Quit: Leaving.)
00:08:25  <CIA-53>node: Ben Noordhuis * r2f8596e / test/simple/test-setproctitle.js :
00:08:25  <CIA-53>node: test: fix bug in setproctitle test
00:08:25  <CIA-53>node: Output from `ps` may contain trailing whitespace, trim it. - http://git.io/AR625Q
00:11:59  <igorzi>bnoordhuis: https://github.com/igorzi/node/commit/c326e1f7cb1dcb771fe609da2f605a7708e716f9
00:12:07  <igorzi>can you pls land?
00:13:17  <CIA-53>node: Igor Zinkovsky * r107b057 / deps/v8/build/common.gypi : build: fix windows build - http://git.io/TUuEAw
00:13:21  <bnoordhuis>igorzi: ^
00:13:51  <igorzi>bnoordhuis: thx again
00:14:18  <bnoordhuis>i live to serve
00:14:54  <bnoordhuis>igorzi: does test/simple/test-net-connect-buffer.js pass for you?
00:15:32  <igorzi>bnoordhuis: nope
00:15:48  <bnoordhuis>good to know it's not just the unices
00:15:52  * bnoordhuisinvestigates
00:21:21  <CIA-53>libuv: Ben Noordhuis * rcdb3601 / AUTHORS : Update AUTHORS and .mailmap - http://git.io/zuXQUQ
00:34:56  <CIA-53>libuv: Ben Noordhuis * r2069a24 / src/unix/stream.c : unix: fix file descriptor leak - http://git.io/Aqih2w
00:45:25  * ericktquit (Quit: erickt)
00:48:40  <isaacs>igorzi: i think i figured out the issue
00:48:49  <isaacs>(re: npm failing randomly)
00:49:08  <igorzi>isaacs: nice.. what was it?
00:49:29  <isaacs>igorzi: well... it looks like having multiple invocations of basic-bsdtar.exe going at the same time causes some problems.
00:49:34  <isaacs>really oddball problems.
00:49:38  <bnoordhuis>[00:59|% 100|+ 264|- 0]: Done <- `make test` on linux, we're getting there
00:49:44  <igorzi>bnoordhuis: one more https://github.com/igorzi/node/commit/8bc95d00d9c17a1aca467fbf33385a6ff5458be1
00:50:11  <isaacs>igorzi: so, if you install something that has a bunch of deps, it'll try to fetch them and unpack them all in parallel, and dies
00:51:03  <CIA-53>node: Igor Zinkovsky * r971c3d9 / src/node.cc : win: remove pthread-win32 init code - http://git.io/y2qR2Q
00:51:04  <igorzi>isaacs: oh.. that's odd that it doesn't like multiple invocations
00:51:23  <isaacs>igorzi: yeah, it's not consistent, either
01:02:25  <igorzi>isaacs: are you planning to "fix" this? (temporarily until you have your js tar ready)?
01:02:31  <igorzi>(by not running in parallel?)
01:08:13  <isaacs>igorzi: yep.
01:08:22  <isaacs>in 2 minutes, you can install 1.0.99, has the fix
01:08:49  <isaacs>igorzi: i'd like to figure out why it is that that fails, though.
01:09:08  <isaacs>if it's a bug in our child_process implementation, then it's going to have to be dealt with eventually.
01:09:54  <isaacs>igorzi: the only clue i have right now is that it seems like having more than 3 unpack operations at the same time, with "pretty big" packages, fails at some random point.
01:10:00  <isaacs>not even always on the same file write operation.
01:10:09  <isaacs>but it's like the child proc returns, and looks like it worked, but didn't.
01:10:40  <isaacs>of course, bsdtar returns the same error code for any "failed to write the file" error, and i'm ignoring that, since almost every package has symlinks, whihc cause it to exit with that code.
01:10:52  <isaacs>it's such a mes.
01:10:54  <isaacs>*mess
01:12:44  <igorzi>isaccs: and i'm assuming that parallel tars aren't trying to write to the same location on disk, correct?
01:32:39  * brsonquit (Quit: leaving)
01:34:06  <isaacs>igorzi: they shouldn't, no
01:34:19  <isaacs>igorzi: no, they can't, because they're bundling all their deps.
01:34:57  <isaacs>oh well. it's taped together for now.
01:34:59  <isaacs>that's fine :)
01:35:28  * isaacsquit (Quit: isaacs)
01:40:48  <CIA-53>node: Ben Noordhuis * r6df574b / (lib/net.js test/simple/test-net-connect-buffer.js): net: properly account multi-byte chars in .bytesWritten - http://git.io/NtPD8A
01:40:49  <CIA-53>node: Ben Noordhuis * r0b92fa0 / lib/net.js : (log message trimmed)
01:40:49  <CIA-53>node: net: fix connect queue bugs
01:40:49  <CIA-53>node: This commit fixes two bugs in the handling of write requests when the connect()
01:40:49  <CIA-53>node: call is still in progress.
01:40:50  <CIA-53>node: 1. The deferred write request's size was counted twice towards `.bytesWritten`.
01:40:50  <CIA-53>node: 2. The callback was not called. After connecting, `Socket.write()` was called
01:40:51  <CIA-53>node: with three arguments (data, encoding, cb) but it ignored the third argument.
02:23:48  * bnoordhuisquit (Ping timeout: 258 seconds)
06:15:03  * ericktjoined
06:27:19  * ericktquit (Quit: erickt)
06:35:34  * Casanjoined
06:36:44  * Casanquit (Client Quit)
06:45:46  * mralephjoined
08:29:49  * dwwwquit (Ping timeout: 240 seconds)
08:45:25  * dwwwjoined
09:15:25  * AvianFlujoined
14:22:13  * mikealjoined
15:17:19  * ericktjoined
15:21:25  * bnoordhuisjoined
16:20:34  <indutny>bnoordhuis: heya
16:20:36  <indutny>how is it going?
16:20:51  <bnoordhuis>indutny: hey
16:20:53  <bnoordhuis>fine. you?
16:20:59  <indutny>fine too
16:21:07  <indutny>going to be in Moscow for next two weeks :)
16:21:14  <indutny>so I'll be in closer TZ to you
16:21:36  <indutny>do you think I can do anything for libuv? /cc ryah
16:22:39  <bnoordhuis>hmm *thinks*
16:23:04  <indutny>ryah: yt? (morning, if so)
16:24:33  <bnoordhuis>i suppose it'd be nice to fix up udp multicast support
16:24:48  <bnoordhuis>there's someone working on that but it seems to have kind of stalled
16:25:20  <indutny>ok
16:25:28  <indutny>do we have issue for that?
16:25:47  <bnoordhuis>https://github.com/joyent/libuv/pull/217
16:26:01  <bnoordhuis>i merged the multicast support patch a while back
16:26:19  <bnoordhuis>but it only works on the unices now (and maybe not even all)
16:26:45  <indutny>ok, sounds good :)
16:26:52  <indutny>looks like I found some other platform issue
16:26:55  <indutny>s/issue/issues
16:27:02  <indutny>so I'll first finish them
16:27:37  <indutny>:)
16:28:32  <bnoordhuis>what kind of issues?
16:29:05  <CIA-53>node: koichik * r68cc173 / (3 files in 3 dirs):
16:29:06  <CIA-53>node: tls: The TLS API is inconsistent with the TCP API
16:29:06  <CIA-53>node: Add 'secureConnect' event to tls.CleartextStream.
16:29:06  <CIA-53>node: Fixes #1467. - http://git.io/AVUaDg
16:29:06  <CIA-53>node: koichik * r86a67f1 / doc/api/tls.markdown : docs: add example of tls - http://git.io/Hes4Dg
16:39:53  <CIA-53>node: koichik * rd6191f5 / (lib/net.js test/simple/test-net-listen-error.js):
16:39:53  <CIA-53>node: net: fix error handling in listen()
16:39:53  <CIA-53>node: Fixes #1894. - http://git.io/LAtGug
16:41:37  <indutny>bnoordhuis: https://github.com/joyent/libuv/issues/142 https://github.com/joyent/libuv/issues/146
16:42:19  <bnoordhuis>indutny: oh, sure
16:44:36  <indutny>forgot about them
16:45:32  <indutny>sorry
17:09:49  * mikealquit (Quit: Leaving.)
17:13:59  * mikealjoined
17:14:57  <bnoordhuis>indutny: what's there to be sorry about?
17:15:13  <indutny>bnoordhuis: for bothering you, while having things to work on :)
17:15:14  <indutny>nevermind
17:16:12  <bnoordhuis>indutny: if i didn't want to be bothered i wouldn't be online (or respond) so don't worry
17:18:16  * piscisaureus_joined
17:18:26  * piscisaureus_changed nick to piscisaureus
17:21:37  <bnoordhuis>piscisaureus: looking forward to the trip?
17:21:46  <piscisaureus>bnoordhuis: yo
17:21:53  <bnoordhuis>yo homie
17:21:55  <piscisaureus>bnoordhuis: yeah
17:21:59  <bnoordhuis>me too
17:22:09  <piscisaureus>bnoordhuis: that's good news
17:22:10  <bnoordhuis>how late do you leave monday?
17:22:15  <bnoordhuis>(too early)
17:22:17  <piscisaureus>*shrug*
17:22:26  <piscisaureus>what time do we leave?
17:22:34  <bnoordhuis>9:50 i think?
17:22:43  <bnoordhuis>i suppose we're on the same train to schiphol
17:23:02  <piscisaureus>bnoordhuis: you are travelling via rotterdam?
17:23:12  <bnoordhuis>yep
17:23:37  <piscisaureus>bnoordhuis: usually I take the sneltrein while I suppose you will take the intercity or fyra
17:23:58  <piscisaureus>bnoordhuis: maybe I'll join you in the hague
17:24:37  <bnoordhuis>okay, cool
17:25:14  <bnoordhuis>hmm, the fyra is more expensive but it's actually a minute slower than the sneltrein
17:25:23  <piscisaureus>bnoordhuis: I think we leave at 11am
17:26:48  <bnoordhuis>piscisaureus: Vertrektijd: 09:50
17:26:56  <bnoordhuis>according to klm.nl
17:27:13  <piscisaureus>bnoordhuis: there are 20 KL 605 flights a day I think
17:27:19  <piscisaureus>KLM ROYAL DUTCH AIRLI 17OCT AMSTERDAM SAN FRANCIS CA 950A 1150A
17:27:19  <piscisaureus>KL 605 MONDAY SCHIPHOL AIRPO INTL
17:27:19  <piscisaureus>N ECONOMY TERMINAL INTL
17:27:19  <piscisaureus> MEAL NON STOP
17:27:19  <piscisaureus> RESERVATION CONFIRMED 11:00
17:27:34  <bnoordhuis>that's why you enter the booking ref :)
17:27:47  <bnoordhuis>11:00 is the flight duration
17:28:07  <rmustacc>950A is the departure time.
17:28:24  <bnoordhuis>piscisaureus: ^
17:29:01  <bnoordhuis>piscisaureus: also, don't forget to do the advance check-in, saves time at the airport
17:29:29  <piscisaureus>aah. these tickets are not the best example of clarity :-)
17:31:54  <rmustacc>How long are you guys around for?
17:32:56  <bnoordhuis>five days, i think?
17:32:59  * ericktquit (Quit: erickt)
17:33:14  <bnoordhuis>yeah, we fly back on sunday
17:33:48  <piscisaureus>bnoordhuis: ok ... dinner
17:33:58  <bnoordhuis>piscisaureus: say hi to the gf from me
17:39:16  * bnoordhuisis off to dinner as well
19:02:22  <indutny>piscisaureus: ;)
19:04:24  <indutny>bnoordhuis: btw, everything fixed here https://github.com/joyent/libuv/pull/203
19:05:30  <indutny>and here too: https://github.com/joyent/libuv/pull/208/files
19:08:43  * indutnyis going to sleep
19:08:44  <indutny>ttyl
19:25:47  <igorzi>bnoordhuis: https://github.com/igorzi/node/commit/a14a492c915d2112fd50d4c3fa9d52b4ea69e46c
19:25:52  <igorzi>please land ---^
19:27:24  * mikealquit (Quit: Leaving.)
19:45:03  * ericktjoined
20:38:37  <dwww>Hi, folks! Sorry for disturbing, but what about raising TCP_NODELAY in TODO? TIA
20:49:38  <piscisaureus>Raising as in bumping priority?
20:49:43  <piscisaureus>hmm
20:49:48  <piscisaureus>patches welcome as we say :-)
20:49:54  <dwww>i know
20:49:58  <piscisaureus>add a github issue, everyone can do that :-)
20:50:10  <dwww>just will have to dive into uv
20:50:37  <piscisaureus>What are you using libuv for btw?
20:50:40  <dwww>do we have socksetopt or such already done
20:50:41  <dwww>?
20:50:50  <dwww>creationix luvit
20:50:53  <piscisaureus>aha
20:50:55  <piscisaureus>erm
20:51:03  <dwww>i came to boundary when i need real-time
20:51:09  <piscisaureus>I'm not doing it immediately (saturday night party)
20:51:11  <dwww>this means NO_DELAY
20:51:15  <piscisaureus>but it'll be there shortly
20:51:22  <dwww>TIA^^^
20:55:57  <CIA-53>node: Igor Zinkovsky * r5ca3dcd / (2 files): test: fix test-child-process-stdin and test-child-process-kill on windows - http://git.io/sKq0mA
20:56:16  <piscisaureus>with shortly I mean within a few weeks at most - but I'm not making any guarantees
20:56:29  <dwww>i see
20:56:47  <dwww>then what could be a workaround?
20:57:43  <dwww>since i do stuff in lua area, i have direct access to *.so (accounting libc.so) via ffi.
20:58:17  <piscisaureus>erm
20:58:27  <dwww>i'd never have disturbed this room if i could do it by myself
20:58:42  <piscisaureus>dwww: well you're welcome to disturb this room
20:59:11  <dwww>direct call to setsockopt of handle->fd with needed parameters gives -1 as status and "
20:59:12  <dwww>unknown" error as stringifyied errno
20:59:24  <piscisaureus>dwww: yes, error mappings may be missing
20:59:27  <dwww>that's why i ever ask here for help
21:00:17  <piscisaureus>dwww: if handle->fd is set (e.g. after the bind call, and on unix) then that should just work
21:00:39  <piscisaureus>dwww: unless you are passing invalid arguments
21:01:44  <dwww>(handle->fd, TCP_NODELAY, (void *)(pointer-to-int-variable-containing-unity), sizeof(int))
21:01:56  <dwww>i'll explore the case deeper
21:02:05  <piscisaureus>dwww: look at errno, not at uv_last_error
21:02:15  <dwww>errno is undefined
21:02:23  <piscisaureus>dwww: eh?
21:02:28  <dwww>whould i include <errno.h>
21:02:31  <dwww>?
21:02:40  <piscisaureus>dwww: "undefined" does not exist in c
21:02:59  <piscisaureus>dwww: what is its value after the setsockopt call?
21:03:11  <dwww>i mean referring errno causes compile error
21:03:12  <dwww>-1
21:03:25  * creationixjoined
21:03:32  <dwww>creationix
21:03:50  <piscisaureus>dwww: it seems that you need to take a c course :-)
21:03:55  <dwww>we're trying to get TCP_NODELAY
21:04:02  <creationix>ahh
21:04:14  <creationix>I came in to ask about how to use uv_udp_*
21:04:17  <dwww>hehe
21:04:18  <creationix>specifically set_membership
21:04:39  <dwww>i'm rather fluent in c
21:05:00  <dwww>just had no time to smoke libuv in
21:05:18  <piscisaureus>dwww: yes, include errno.h
21:05:25  <dwww>k
21:07:46  <piscisaureus>dwww: also, take care to try setsockopt *after* calling uv_bind_xxx
21:07:58  <piscisaureus>dwww: since actual sockets are created lazily
21:08:21  <piscisaureus>dwww: that's all the pointers I can give you. gtg now. if you can't get around, we'll add this feature in the near future
21:08:52  <dwww>thanks for the kludges
21:09:11  <dwww>i'll feedback once succeded
21:23:20  * brsonjoined
21:45:00  * dwwwquit (Quit: Leaving.)
21:51:21  * Casanjoined
22:15:08  * mralephquit (Quit: Leaving.)
23:08:57  * bnoordhuisquit (Read error: Operation timed out)
23:23:11  * ericktquit (Quit: erickt)
23:41:31  <creationix>ok, so anyone around who knows how to use the udp stuff in libuv?