00:03:32  * piscisaureus_quit (Ping timeout: 250 seconds)
00:11:57  * brsonquit (Ping timeout: 272 seconds)
00:23:08  * brsonjoined
00:30:03  * avalanche123quit (Quit: Computer has gone to sleep.)
00:34:45  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
00:38:03  * avalanche123joined
00:53:12  * dshawquit (Ping timeout: 264 seconds)
01:11:28  * TheJHquit (Ping timeout: 244 seconds)
01:33:52  * davisp-changed nick to davisp
01:33:53  * davispquit (Changing host)
01:33:53  * davispjoined
01:35:31  * bnoordhuisquit (Ping timeout: 240 seconds)
02:05:56  * beachdogjoined
02:11:57  * beachdog_joined
02:15:27  * brsonquit (Ping timeout: 272 seconds)
02:34:23  * dylangquit (Quit: dylang)
02:44:02  * dylangjoined
02:45:22  * dshaw_joined
03:01:23  * dylangquit (Quit: dylang)
03:12:37  * beachdog_quit (Remote host closed the connection)
03:30:27  * mmaleckiquit (Ping timeout: 244 seconds)
03:32:28  * c4milojoined
03:34:51  * beachdogquit (Remote host closed the connection)
03:38:47  * c4miloquit (Remote host closed the connection)
03:42:19  * brsonjoined
03:46:40  * avalanche123quit (Quit: Computer has gone to sleep.)
03:48:25  * avalanche123joined
03:52:09  * jgautierquit (Ping timeout: 245 seconds)
04:03:23  * theColejoined
04:09:01  * avalanche123quit (Quit: Computer has gone to sleep.)
05:28:56  * ProLoser|Macjoined
05:28:56  * milanijoined
05:29:03  <ProLoser|Mac>hallo
05:29:18  <ProLoser|Mac>anyone around?
05:30:04  * ProLoser|Macpart
05:40:07  * milaniquit (Quit: Ex-Chat)
06:06:10  * AvianFlu_joined
06:06:36  * AvianFluquit (Disconnected by services)
06:06:37  * AvianFlu_changed nick to AvianFlu
06:08:55  * theColequit (Quit: theCole)
06:24:39  * dshaw_quit (Quit: Leaving.)
06:44:39  * stephankquit (Quit: *Poof!*)
06:56:03  * AvianFluchanged nick to AvianusAsleepus
07:02:36  * theColejoined
07:04:23  * `3rdEdenjoined
07:15:53  * txdvquit (Read error: Connection reset by peer)
07:18:39  * txdvjoined
07:26:43  * mmaleckijoined
07:30:59  * mmaleckiquit (Ping timeout: 246 seconds)
07:31:54  * rendarjoined
07:35:07  * indexzeroquit (Quit: indexzero)
07:44:23  * brsonquit (Ping timeout: 246 seconds)
08:07:07  * dshaw_joined
08:14:35  * loladirojoined
08:17:10  * hzjoined
08:31:33  * stagas_joined
08:36:35  * AndreasMadsenjoined
08:41:55  * mmaleckijoined
08:55:42  * loladiroquit (Quit: loladiro)
09:11:32  * loladirojoined
09:13:53  * charliesomejoined
09:19:28  * piscisaureus_joined
09:52:02  * piscisaureus_quit (Ping timeout: 246 seconds)
10:01:03  * stagas_quit (Ping timeout: 244 seconds)
10:32:18  <indutny>hey people
10:33:26  <mmalecki>hello!
10:36:20  * loladiroquit (Quit: loladiro)
10:37:32  * piscisaureus_joined
10:41:33  <indutny>mmalecki: interesting question!
10:41:39  <indutny>how to broadcast UDP packets
10:41:45  <indutny>if I can't get network mask in node.js :D
10:42:40  <mmalecki>indutny: that indeed is an interesting question
10:43:59  <indutny>and why do I need a API for broadcasting in node if I can't use them
10:44:36  * saschakbjoined
10:45:18  <indutny>piscisaureus_: ping
10:46:12  * saschakbpart ("Verlassend")
10:47:02  * charliesomequit (Ping timeout: 244 seconds)
10:57:08  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
11:05:06  * beachdogjoined
11:06:44  * dylangjoined
11:14:22  * beachdogquit (Remote host closed the connection)
11:25:10  * bnoordhuisjoined
11:27:17  * charliesomejoined
11:36:57  * piscisaureus_joined
11:41:36  <indutny>bnoordhuis: hey man
11:41:42  <bnoordhuis>indutny: ho man
11:41:53  <indutny>bnoordhuis: what do you think about exposing network mask in uv's get interfaces
11:42:11  <bnoordhuis>indutny: no
11:42:22  <bnoordhuis>libuv is accumulating enough cruft as it is
11:42:28  <indutny>oh
11:42:36  <indutny>how would one send broadcast packets then?
11:42:48  <bnoordhuis>i just shot down marak on the nodejs-dev list regarding routing tables
11:43:11  <mmalecki>bnoordhuis: link to that topic?
11:43:19  <bnoordhuis>mmalecki: groups.google.com
11:43:35  <bnoordhuis>mmalecki: (iow, i don't know - i only use email)
11:46:36  <mmalecki>bnoordhuis: I see.
11:47:18  <mmalecki>I mean, I'm quite +1 on adding any way to gather OS information into the core
11:47:30  <bnoordhuis>noted. i'm not
11:47:58  <mmalecki>the reason for that is that we're deploying our monitoring software to over 3 thousand servers, compiling native addons on those is really error prone
11:49:45  <bnoordhuis>mmalecki: precompile them
11:50:14  <indutny>bnoordhuis: build farms!!!!
11:50:24  <mmalecki>bnoordhuis: won't work on some linux machines, but indeed, this is what we're doing on smartos
11:50:25  <indutny>bnoordhuis: so lets remove dgram support then
11:50:29  <bnoordhuis>indutny: okay
11:50:33  <indutny>bnoordhuis: and tcp
11:50:37  <indutny>bnoordhuis: just in case
11:50:42  <mmalecki>yeah, I never use tcp anyway
11:50:45  <bnoordhuis>i know you're joking about tcp
11:50:46  <indutny>me too
11:50:48  <indutny>fuck that http
11:50:48  <mmalecki>I'd be +1 on removing it
11:50:51  <bnoordhuis>but i could live with moving dgram out of core
11:50:59  <mmalecki>it's lots of useless cruft
11:51:01  <indutny>bnoordhuis: yeah, I'm ok with that
11:51:12  <indutny>bnoordhuis: but it should be somewhere, not just out of core
11:51:14  <mmalecki>bnoordhuis: <hint>dns</hint>
11:51:33  <indutny>so, speaking of wierd stuff
11:51:42  <indutny>how could one get router address?
11:51:43  <mmalecki>I mean, I consider udp one of those crucial network protocols
11:51:46  <indutny>s/router/gateway/
11:51:54  <indutny>mmalecki: well, vock really relies on it
11:52:01  <indutny>and almost every mature online game too
11:52:02  <indutny>brb
11:52:06  <bnoordhuis>mmalecki: it is but not many node users actually use it
11:52:08  <mmalecki>indutny: our dns server relies on it :)
11:52:16  <bnoordhuis>mmalecki: re dns, we don't use the dgram module for that
11:52:30  <mmalecki>bnoordhuis: I mean, DNS servers
11:53:09  <bnoordhuis>solved with one more line in your package.json
11:53:46  <mmalecki>well, okay
11:53:53  <mmalecki>but still, half of internet uses udp
11:54:03  <mmalecki>and node is supposed to support all crucial network protocols
11:54:19  <piscisaureus_>I am actually okay with adding more network cruft
11:54:33  <mmalecki>piscisaureus_: gopher support?
11:54:38  <piscisaureus_>I'd be fine with adding gateway, mac, dns server to os.networkInterfaces()
11:54:51  <mmalecki>piscisaureus_: ++
11:54:51  <kohai>piscisaureus_ has 15 beers
11:54:59  <bnoordhuis>piscisaureus_: well, i'm not
11:55:00  <piscisaureus_>but totalMem(), totalSwap(), disks() <-- no
11:55:45  <piscisaureus_>The amount of disagreement is growing in node-core
11:55:51  <bnoordhuis>i wouldn't mind writing an add-on if there's enough demand
11:56:19  <bnoordhuis>and if it turns out that it's used by hundreds or thousands of people, we can move it into core
11:57:19  <piscisaureus_>we have very little insight in how manu people use a modul.e
11:57:45  <bnoordhuis>you can tell by the number of bug reports and feature requests you get
11:58:04  <bnoordhuis>i know that node-iconv is used a lot. node-dgram-unix? not so much
11:59:14  <piscisaureus_>node-dgram-unix only works on unix
11:59:19  <piscisaureus_>so nobody can use it
12:00:12  <piscisaureus_>bnoordhuis: I think udp is important. It increases node's credibility
12:00:39  <piscisaureus_>oh wow
12:00:48  <bnoordhuis>piscisaureus_: moving it into a module would increase node's credibility as a platform you can extend
12:00:51  <piscisaureus_>bnoordhuis: that iconv thing actually does work on this windows thing
12:00:58  <bnoordhuis>yes :)
12:01:13  <bnoordhuis>the beauty of open source
12:03:20  <mmalecki>meh, I guess I'll get some food instead of nerdfighting about udp :)
12:04:11  <piscisaureus_>I like most stuff that is in core right now
12:04:40  <piscisaureus_>except maybe cares
12:04:51  <mmalecki>so do I. it gives me that warm and fuzzy feeling
12:05:14  <piscisaureus_>although, if I had to do it again, there would be no eventEmitters :-)
12:05:16  <mmalecki>with a bit of chill from being close to bare metal
12:06:31  <piscisaureus_>let's move crypto/ssl out of core
12:07:29  * mmaleckichanged nick to mmalecki[food]
12:15:33  * piscisaureus__joined
12:16:13  * piscisaureus__changed nick to piscisaureus
12:16:37  <piscisaureus>bnoordhuis: rather than throwing dns out, it'd rather throw cares out and move to a pure-js dns impl
12:16:55  <piscisaureus>bnoordhuis: although we'd have to add uv_dns_servers()
12:18:11  <bnoordhuis>piscisaureus: agreed
12:18:25  <piscisaureus>sounds like a plan
12:18:25  <bnoordhuis>make it so
12:18:36  <piscisaureus>I think tjfontaine was working on a dns implementation
12:18:41  * piscisaureus_quit (Ping timeout: 246 seconds)
12:18:42  * charliesomequit (Quit: Textual IRC Client: www.textualapp.com)
12:18:43  <piscisaureus>I'm going to throw cares out of libuv
12:18:45  <piscisaureus>NOW
12:19:42  <bnoordhuis>just make sure you don't break anything
12:29:37  * stagas_joined
12:31:16  <indutny>bnoordhuis: so, have you agreed with adding gateway subnet mask to node?
12:44:16  <bnoordhuis>indutny: no
12:44:38  <bnoordhuis>what's up with that global notice btw?
12:45:32  <indutny>рур
12:45:32  <indutny>heh
12:45:33  <indutny>dunno
12:56:50  * dylangquit (Quit: dylang)
13:00:27  * stagas_quit (Ping timeout: 250 seconds)
13:25:25  * beachdogjoined
13:42:37  * beachdogquit (Remote host closed the connection)
13:44:42  * bradleymeckjoined
13:57:01  <tjfontaine>I can't tell if that was a serious conversation about dgram
13:59:29  <ArmyOfBruce>when you talk about moving to a pure-JS version, is that a sign that those of us using libuv from another language aren't equal status users?
13:59:38  <ArmyOfBruce>if so, find ... but would be nice to be clear about that.
14:02:51  * c4milojoined
14:04:58  <piscisaureus>ArmyOfBruce: c-ares will be removed but integrating c-ares yourself will be trivial
14:05:11  <piscisaureus>ArmyOfBruce: I am prepping a repo and a gist with glue code
14:05:21  <piscisaureus>Any sed experts here? I never get that right///
14:05:23  <piscisaureus>sed /config_/config\\//
14:05:46  <piscisaureus>How do I replace "src/config_bla" by "config/bla" ?
14:08:54  <tjfontaine>s#src/config_bla#config/bla#
14:09:17  <tjfontaine>you can use just about anything to indicate barriers for replacement, so long as you match them
14:10:03  <piscisaureus>yeah
14:10:19  <piscisaureus>so that outputs "onfig//" for me :-)
14:10:30  <piscisaureus>I am now trying a perl one liner
14:10:50  <tjfontaine>o0
14:11:16  <tjfontaine>echo "src/config_blah" | sed -e s#src/config_blah#config/blah#
14:11:17  <tjfontaine>config/blah
14:11:56  <piscisaureus>Ah I see
14:12:00  <piscisaureus>I was missing the 's'
14:12:05  <piscisaureus>So I switched to
14:12:05  <piscisaureus>perl -pe 's/zzzzsrc\/config_/config\//'
14:12:11  <piscisaureus>which also works :-)
14:12:38  <tjfontaine>heh
14:14:39  * AndreasMadsenquit (Remote host closed the connection)
14:16:06  <bradleymeck>how do you force dest-cpu in gyp's configure
14:16:13  * xaqjoined
14:16:32  <piscisaureus>umm, I don't know
14:16:42  <piscisaureus>although I've ssen mentions of it on the mailing list
14:17:20  <bradleymeck>SunOS seems to ignore --dest-cpu...
14:24:13  <indutny>bradleymeck: hey man
14:24:16  <indutny>oops
14:24:18  <indutny>bnoordhuis: ping
14:24:19  <indutny>hey
14:24:47  <indutny>bnoordhuis: https://github.com/joyent/node/pull/3661/files#r1292292
14:24:57  <indutny>bnoordhuis: what is the signature of non-copying constructor?
14:25:27  <indutny>bnoordhuis: using with free callback?
14:26:28  <indutny>yeah, looks like so
14:26:29  <indutny>nvm
14:40:40  * xaqquit (Remote host closed the connection)
14:49:25  * AvianusAsleepuschanged nick to AvianFlu
14:52:06  * xaqjoined
14:54:02  * bnoordhuisquit (*.net *.split)
14:54:02  * rendarquit (*.net *.split)
14:54:02  * stagasquit (*.net *.split)
14:54:02  * chiltsquit (*.net *.split)
14:54:03  * einarosquit (*.net *.split)
14:55:27  * c4milo_joined
14:56:38  * bnoordhuisjoined
14:56:38  * rendarjoined
14:56:38  * stagasjoined
14:56:38  * chiltsjoined
14:56:38  * einarosjoined
14:56:43  * c4miloquit (Ping timeout: 263 seconds)
15:11:45  * theColequit (Quit: theCole)
15:11:49  * bradleymeck_joined
15:12:20  * theColejoined
15:15:42  * bradleymeckquit (Ping timeout: 244 seconds)
15:15:42  * bradleymeck_changed nick to bradleymeck
15:15:56  * c4milojoined
15:16:38  * c4milo_quit (Ping timeout: 246 seconds)
15:20:17  * mmalecki[food]changed nick to mmalecki
15:20:47  * c4miloquit (Ping timeout: 272 seconds)
15:24:19  <CIA-134>node: isaacs master * r0414e14 / doc/api/util.markdown : doc: util.pump is deprecated. Use Stream.pipe - https://github.com/joyent/node/commit/0414e14702af7258a9197f6b979e32261f96f910
15:27:01  * stagasquit (Ping timeout: 240 seconds)
15:28:01  <isaacs>bnoordhuis: https://github.com/isaacs/node/commit/26573da1a883f8ab8f9fb5a9273c68a51c225f9c
15:31:53  * AvianFluchanged nick to Away-vianFlu
15:32:18  * `3rdEdenquit (Quit: fooding <3)
15:32:32  * TooTallNatejoined
15:37:06  <indutny>bnoordhuis: yt?
15:37:11  <bnoordhuis>indutny: ih
15:37:17  <indutny>bnoordhuis: so, about aligned loads
15:37:24  <indutny>bnoordhuis: how should it be done?
15:37:29  <indutny>https://github.com/joyent/node/pull/3661/files#r1292348
15:37:40  <indutny>should I do data_[3] * 256 + data_[4] ?
15:37:42  <indutny>will that work?
15:37:53  <bnoordhuis>indutny: yes
15:38:05  <bnoordhuis>but make sure that data_ is unsigned char[]
15:38:10  <indutny>uint8_t
15:38:15  <bnoordhuis>good
15:39:37  <bnoordhuis>isaacs: reviewed
15:41:59  * Away-vianFluquit (Read error: Connection reset by peer)
15:42:08  * dapjoined
15:42:09  <CIA-134>node: isaacs master * r168a555 / (5 files in 5 dirs): fs: fix naming of truncate/ftruncate functions - https://github.com/joyent/node/commit/168a5557803b40b4930c3b4f9587b67a6d566b13
15:42:16  * AvianFlujoined
15:42:29  <isaacs>bnoordhuis: of course you realize the `else if` is semantically equivalent, since len cannot be both 'function' and 'undefined'
15:42:35  <isaacs>bnoordhuis: but for you, i will paint all the bikesheds.
15:45:31  * chobi_e_changed nick to chobi_e
15:47:10  * c4milojoined
15:49:10  * ericktjoined
15:50:02  * dshaw_quit (Quit: Leaving.)
15:57:24  * dshaw_joined
15:57:56  * chobi_echanged nick to chobi_e_
15:58:52  * c4milo_joined
15:59:42  * AvianFlu_joined
16:01:34  * c4miloquit (Ping timeout: 250 seconds)
16:02:22  * AvianFluquit (Ping timeout: 246 seconds)
16:02:23  * AvianFlu_changed nick to AvianFlu
16:02:47  <indutny>bnoordhuis: fixed every mentioned issue in my PR
16:02:53  <indutny>bnoordhuis: https://github.com/indutny/node/commit/0488bd70db03f411d942e8a47e2b7c16d113e812
16:04:39  <piscisaureus>creationix: hey
16:04:54  <piscisaureus>creationix: is there any way to append to an existing file w/ vfs
16:04:56  <piscisaureus>?
16:05:30  <indutny>brb
16:12:36  <bnoordhuis>isaacs: in purple please
16:16:39  * stephankjoined
16:17:00  * AvianFlu_joined
16:20:19  * AvianFluquit (Ping timeout: 272 seconds)
16:20:20  * AvianFlu_changed nick to AvianFlu
16:21:43  <indutny>bnoordhuis: huh?
16:22:36  * stagasjoined
16:22:56  <creationix>piscisaureus, nope
16:23:16  <creationix>piscisaureus, I wasn't sure how portable or useful that would be
16:23:27  <piscisaureus>creationix: hmm, ok. I need it :-(
16:23:35  <creationix>for what backend?
16:23:41  <piscisaureus>creationix: ssh
16:23:47  <piscisaureus>creationix: obviously I can work around it...
16:23:59  <piscisaureus>creationix: but append files are nice, writes are atomic etc
16:24:14  * isaacstopic: Liberty. Unity. Velociraptors. http://piscisaureus.no.de/libuv/latest
16:24:15  <creationix>can node
16:24:27  <creationix>can node's createWriteStream open in append mode and write to it?
16:24:32  <isaacs>creationix: yep
16:24:45  <isaacs>fs.creatWriteStream(path, { flags: 'a' })
16:24:47  <isaacs>iirc
16:24:57  <isaacs>s/creat/create/
16:25:08  <indutny>bnoordhuis: does my patch lgty now?
16:25:48  <creationix>piscisaureus, ok, so the reason I don't open over the existing file is because it could be a symlink and bypass security checks
16:25:57  <creationix>but I guess I could put in a lstat to check that
16:26:07  <piscisaureus>creationix: racy :-)
16:26:17  <creationix>yeah, I wish there was flstat
16:26:23  <piscisaureus>that'd be better indeed
16:26:30  <piscisaureus>only on windows :-p
16:26:59  <creationix>right now I create a temp file using exclusive mode and then rename over the original path
16:27:08  <creationix>how would I add append support that's still secure
16:27:36  * brsonjoined
16:27:40  <piscisaureus>creationix: is it possible to see the path for an open fd
16:27:50  <piscisaureus>creationix: in that case you would open it and then check the actual path
16:28:12  <isaacs>piscisaureus: flstat would have no meaning on systems that don't support O_SYMLINK
16:28:29  <isaacs>piscisaureus: since opening a symlink on any non-Darwin unix === opening the target
16:28:50  <creationix>piscisaureus, oh, so fs.open and then fs.fstat to get the path?
16:29:02  <piscisaureus>creationix: I don't think fstat gives you the path
16:29:02  <creationix>I don't think node exposes that
16:29:03  <isaacs>piscisaureus: and flstat on Darwin would just be the same as fstat() after you open the symlink with O_SYMLINK
16:29:11  * AvianFlu_joined
16:29:22  <isaacs>creationix: no, there's no way to get teh path from the fd
16:29:29  <isaacs>or the path from the stat
16:29:31  <piscisaureus>so actually what we need is O_SYMLINK
16:29:32  <isaacs>you get the dev+ino
16:29:40  <isaacs>the path is a link to dev+ino, not the other way around
16:29:58  <creationix>piscisaureus, well, for your case, you can just use vfs.extend and write custom node code (bypassing all security)
16:30:17  <creationix>but I would like a portable and secure append in vfs itself if possible
16:30:23  <piscisaureus>creationix: so how can you be sure that moving a file is safe?
16:30:42  <piscisaureus>creationix: because people might quickly symlink the dir that the target is in
16:31:03  <creationix>the temp file is in the same dir as the target
16:31:08  <creationix>not sure if that helps though
16:31:16  <piscisaureus>creationix: so, what if the dir was a symlink?
16:31:28  <creationix>rename is path based, hmm
16:31:41  <creationix>so basically, anything path based is racy
16:31:45  <piscisaureus>yes
16:31:51  <piscisaureus>some unices have openat() etc
16:31:54  <isaacs>why there is no flink in posix is beyond me.
16:31:58  <piscisaureus>to make it slightly better
16:31:59  <isaacs>that is so obnoxious, imo
16:32:12  <piscisaureus>yeah
16:32:13  <creationix>do we even use that feature of vfs? (running several clients in the same root node process)
16:32:16  <piscisaureus>and frealpath :-)
16:32:34  <creationix>vfs-ssh runs as the real user and so the linux container does all the security
16:32:48  * AvianFluquit (Ping timeout: 264 seconds)
16:32:49  * AvianFlu_changed nick to AvianFlu
16:32:59  <piscisaureus>creationix: I think we did on joyent, but nowadays it's all gears I think
16:33:21  <piscisaureus>creationix: yeah the vfs process should be properly sandboxed
16:33:24  <creationix>I'm wondering if it's a feature worth keeping
16:33:30  <creationix>that level of security is always hard
16:33:38  * devaholicquit (Ping timeout: 250 seconds)
16:33:38  <creationix>and doesn't work on non-unix filesystems anyway
16:33:44  <creationix>since I assume posix semantics
16:34:36  <piscisaureus>creationix: you probably want to discuss this with Knuth
16:34:37  <creationix>removing the user-space sandboxing would make vfs-local about 1/3 the size and a little faster
16:34:47  <creationix>yep
16:34:51  <isaacs>creationix: to do appends, can you first copy the file to a temp, then append to that, and then do the normal atomic mv?
16:35:09  <piscisaureus>isaacs: that's what I would need to do. But it's slow(er)
16:35:11  <creationix>isaacs, that kinda defeats the purpose of an append I think
16:35:12  <isaacs>piscisaureus: frealpath is impossible, you know that
16:35:13  <piscisaureus>and more tedious
16:35:27  <piscisaureus>isaacs: it's not. Windows has it :-p
16:35:33  <isaacs>piscisaureus: does windows have hard links?
16:35:35  <creationix>I'm going to see if my team is ok with removing the user-space sandbox
16:35:40  <piscisaureus>isaacs: it has
16:35:41  <creationix>that's the best solution imo
16:36:01  <isaacs>piscisaureus: do windows handles know what path you opened?
16:36:12  <piscisaureus>isaacs: http://msdn.microsoft.com/en-us/library/windows/desktop/aa364962%28v=vs.85%29.aspx
16:36:27  <isaacs>piscisaureus: i don't get how it can be possible to get the path from an integer that points to a inode, since that inode can be at multiple paths
16:36:41  <piscisaureus>isaacs: sure, but atleast it can give you one path
16:37:18  <isaacs>hm.
16:38:10  <isaacs>i rather like the elegance of path as a persistent link to an inode, fd as a temporary link to an inode, and inodes as the real "file" thing
16:38:58  <isaacs>the lack of flink breaks the elegance.
16:39:22  <isaacs>because then you could have true temporary files and single atomic writes.
16:39:30  <piscisaureus>symlinks just suck
16:39:32  <piscisaureus>let's admit it
16:39:38  <creationix>symlinks are just path tricks
16:39:51  <isaacs>piscisaureus: well, they were underspecified. that's what sucks.
16:39:54  <creationix>maybe they should be treated as text-files and not magical files
16:40:05  <isaacs>creationix: on os x, that's basically what they are.
16:40:10  * bradleymeckchanged nick to bradleymeck[lunc
16:40:13  <isaacs>creationix: text files with a special mode flag.
16:40:26  <isaacs>and the contents of the file is the target.
16:40:33  <creationix>right, but it's the magic mode that makes them so useful and dangerous
16:40:56  <creationix>linux is the same afaik
16:40:59  <isaacs>i mean, clearly there is a valid use case for putting shortcuts between things on the fs
16:41:02  * AvianFlu_joined
16:41:06  <isaacs>creationix: but on linux, you can't open them
16:41:18  <creationix>not using normal open
16:41:20  <creationix>right
16:41:21  * bradleymeck[luncquit (Quit: bradleymeck[lunc)
16:41:24  <creationix>readlink or something
16:43:31  * AvianFluquit (Ping timeout: 240 seconds)
16:43:33  * AvianFlu_changed nick to AvianFlu
16:45:29  <creationix>piscisaureus, knuth team agrees
16:45:50  <creationix>I hope nobody is depending on the sandbox on vfs-local (it is open source after all)
16:46:37  * ericktquit (Quit: erickt)
16:51:20  * ericktjoined
16:52:09  * mmaleckichanged nick to mmalecki[away]
16:53:24  <creationix>piscisaureus, ok, when I'm done with the new vfs version, you can just use node's "a" flag and it will work as expected. createReadStream options are passed through
16:54:47  <piscisaureus>creationix: cool
16:55:11  * cocacolajoined
17:11:19  * cocacolaquit (Quit: Page closed)
17:19:22  <piscisaureus>our common.gypi's are full of cargo cult
17:21:55  <creationix>piscisaureus, I think I agree
17:22:15  <creationix>I tried to remove some of it my projects, but never had time to properly learn gyp
17:22:25  <piscisaureus>neither do I
17:22:34  <piscisaureus>but some pruning wouldn't hurt
17:22:55  <piscisaureus>just remove stuff until it breaks :-)
17:27:05  <TooTallNate>does anyone have any final objections? https://github.com/joyent/node/pull/3824
17:27:28  <TooTallNate>isaacs: bnoordhuis: piscisaureus: indutny: ^
17:41:28  * joshthecoderjoined
17:49:10  <indutny>TooTallNate: dunno
17:49:17  <indutny>that's too complicated
17:49:26  <indutny>and I've never used it
17:49:31  <TooTallNate>which part?
17:50:06  <indutny>make platform
17:50:07  * c4milojoined
17:50:30  * c4milo_quit (Read error: Connection reset by peer)
17:51:00  <TooTallNate>you mean `make binary`?
17:55:56  * c4milo_joined
17:57:53  * brson_joined
17:58:01  * brsonquit (Read error: Connection reset by peer)
17:58:19  * c4miloquit (Ping timeout: 246 seconds)
18:00:14  * bradleymeckjoined
18:04:45  * AndreasMadsenjoined
18:07:28  <piscisaureus>https://github.com/piscisaureus/cares/blob/master/cares.gyp <-- cares removal step 1
18:17:59  <isaacs>TooTallNate: reading through it again
18:19:35  <indutny>piscisaureus: btw
18:19:45  <indutny>piscisaureus: wanna pick up opus.gyp and speex.gyp ?
18:20:04  <mscdex>opus.js!
18:20:57  <indutny>heh
18:21:04  <indutny>that's up to you
18:21:21  <mscdex>there's already a speex.js
18:21:57  <mscdex>i think the author used emscripten though
18:22:13  <mscdex>it's compatible with aurora.js
18:22:31  * theColequit (Quit: theCole)
18:22:44  <isaacs>bikeshed request: https://gist.github.com/3277318
18:22:46  <mscdex>opus looks very interesting though
18:23:16  <isaacs>i think that url.format() should favor the more "specific" fields over the composite fields
18:23:38  <isaacs>ie, hostname+port should trump the host field, in the same way that pathname+search overrides the path field
18:28:11  * avalanche123joined
18:28:33  <isaacs>TooTallNate: Add "release-only", "tar", and "binary" to the .PHONY line at the bottom of the Makefile
18:29:04  <TooTallNate>isaacs: ohh right, forgot about that, one sec
18:30:22  <isaacs>TooTallNate: also, either squash into a few smaller commits, or merge into v0.8 with --no-ff (or both), so that we can easily keep the whole feature grouped nicely
18:31:02  <isaacs>TooTallNate: the shebang hack kind of is irksome, but we've been around and around that issue, and i don't think there's really a better solution
18:31:20  <TooTallNate>me either unfortuantely
18:31:29  <TooTallNate>i was hoping ben has some trick up his sleeve :p
18:31:33  * mmalecki1away]joined
18:32:04  <isaacs> // >/dev/null; is particularly devious.
18:32:20  <isaacs>beautiful, even.
18:32:36  <isaacs>like a beautiful viper you'd be afraid of.
18:33:27  <isaacs>but yes, let's do this for 0.8.6
18:33:29  <isaacs>that'll be rad
18:33:39  * beachdogjoined
18:34:00  <isaacs>TooTallNate: where did you build the sunos one? on the smartosdrone server?
18:34:09  <TooTallNate>isaacs: umcats
18:34:37  <isaacs>kewl
18:34:42  <isaacs>you knwo there's a bigger one?
18:34:48  <isaacs>umcats is kind of a wimp
18:35:07  <isaacs>Host smartos HostName 165.225.128.184 ForwardAgent yes
18:35:16  <isaacs>TooTallNate: make -j32 on that biatch
18:35:33  <TooTallNate>holy crap
18:35:33  <TooTallNate>oh
18:35:34  <TooTallNate>ok
18:35:48  <TooTallNate>isaacs: oh btw my pub key fails for smartos64drone
18:36:04  <isaacs>it has 8GB of ram, and 24 virtual cores, but actually it's 32
18:36:08  <isaacs>orly?
18:36:16  <isaacs>TooTallNate: you logging in as root?
18:36:41  <isaacs>oh, smartos64, yeah, that's gone
18:36:45  <isaacs>turns out there's actually no difference
18:36:54  <TooTallNate>oh ok
18:36:54  <TooTallNate>:p
18:36:56  <mscdex>i'm starting to think it's going to be easier to just write an ssh library in plain js
18:37:01  <mscdex>:\
18:37:08  <TooTallNate>mscdex: i've been dreaming about that ;)
18:37:11  <isaacs>mscdex: DO IT! :) :) :)
18:37:16  <AvianFlu>mscdex: talk to bradleymeck
18:37:21  <mscdex>i kind of started already
18:37:34  <mscdex>libssh2 seems good, but it's got some weird things about it
18:37:45  * KiNgMaRquit (Excess Flood)
18:38:17  <TooTallNate>mscdex: apparently something about the event loops being hard to merge or something?
18:38:17  <mscdex>like a foo_free() that can somehow return EAGAIN, when you'd think it's just freeing memory
18:38:28  <mscdex>no, libssh2 can do async
18:38:46  <mscdex>it returns EAGAIN on stuff that would block
18:38:50  <bradleymeck>mscdex: i have an old partiallyworking ssh lying around, but if you want to make one in js that would be nice
18:38:56  * avalanche123quit (Ping timeout: 246 seconds)
18:39:17  <bradleymeck>libssh2 gets kind of wonky when you want to do things like have keys not tied to users on the host
18:39:35  * KiNgMaRjoined
18:39:59  * mmalecki[away]quit (Remote host closed the connection)
18:40:11  * c4milo_quit (Read error: Connection reset by peer)
18:40:21  <mscdex>besides that libuv bug was also bugging me
18:40:35  * c4milojoined
18:41:04  <isaacs>TooTallNate: so, yeah, land that binary stuff. if you're feeling ambitious, you can also update the tools/email-footer.md file to include the urls to the built tarballs
18:41:14  <isaacs>TooTallNate: and we can make that a regular part of teh release.
18:41:24  <TooTallNate>i can do that
18:41:33  <isaacs>TooTallNate: where'd you build the linux one? centosdrone or ubuntudrone or somewhere else?
18:41:52  <TooTallNate>my (about to be decommissioned) server at home :D
18:42:01  <TooTallNate>but i'll use ubuntudrone for the releases
18:42:12  <TooTallNate>isaacs: also, centosdrone seems down
18:43:27  <bradleymeck>TooTallNate: does node-gyp ever mess with LDFLAGS or does that get passed on?
18:43:48  <TooTallNate>bradleymeck: it never touched LDFLAGS
18:44:00  <TooTallNate>but they should be passed to gyp and make
18:44:02  <bradleymeck>k~ must be something else then
18:46:00  <isaacs>TooTallNate:
18:46:01  <isaacs>$ ssh centosdrone 'hostname;whoami'
18:46:01  <isaacs>30488f3c-def1-49e3-8a01-6eca90842370
18:46:01  <isaacs>root
18:46:16  <isaacs> HostName 165.225.130.68
18:46:25  <isaacs>TooTallNate: centos is that ip^
18:46:29  <TooTallNate>hmm, my ip was wrong
18:46:45  * c4milo_joined
18:46:48  <TooTallNate>ok works for me now :)
18:47:15  * avalanche123joined
18:47:57  <mscdex>bradleymeck: was your ssh thing pure js or was it an addon?
18:48:57  * c4miloquit (Ping timeout: 246 seconds)
18:49:47  * brson_quit (Ping timeout: 246 seconds)
18:50:01  <bradleymeck>mscdex pure js, needed to have non-standard key negotiation for the little use case i was doing
18:50:08  <bradleymeck>i need to do a full rewrite though
18:50:14  <bradleymeck>which i never have time for
18:50:54  * c4milo_quit (Read error: Connection reset by peer)
18:51:02  * c4milojoined
18:51:24  <mscdex>bradleymeck: i'd be interested in taking a look at what you have if you don't mind
18:51:50  <bradleymeck>mscdex, just let me know where to email it, its on a diff machine
18:53:18  * EhevuTovjoined
18:55:21  <TooTallNate>isaacs: is there a file i should update for the md5 checksums?
18:56:56  * brsonjoined
18:57:00  <isaacs>TooTallNate: no, i generate those manually
18:57:10  <isaacs>TooTallNate: and i guess i'll just gpg sign the SHASUMS.txt file
18:58:01  <isaacs>TooTallNate: it'd be kinda nice, i guess, if the tarball included a shasum of all the binaries it includes, but that's kind of excessive, probably
18:58:02  * pfox__quit (Ping timeout: 240 seconds)
18:58:19  <isaacs>if any of them are different, then the sum of the tarball will be different
18:58:43  <isaacs>in fact, that's better, because it also shasums over the mode and time info
18:59:59  * EhevuTov_joined
19:02:57  * EhevuTovquit (Ping timeout: 246 seconds)
19:08:46  <TooTallNate>isaacs: are you gonna give me scp access to that dist/ dir, or what?
19:09:40  * theColejoined
19:12:44  * lohkeyjoined
19:16:40  * theColequit (Quit: theCole)
19:24:50  <TooTallNate>ok; merging...
19:25:18  <isaacs>TooTallNate: if you're volunteering to help with builds, then fuck yeah
19:25:40  <TooTallNate>well… ya i am :D
19:25:42  <TooTallNate>you work too hard isaacs
19:26:08  <tjfontaine>wait, isn't that why they pay him the big bucks?
19:26:21  * stagas_joined
19:27:00  <CIA-134>node: Nathan Rajlich v0.8 * ra23df77 / (Makefile tools/email-footer.md tools/install.py): Merge branch 'binary-tarballs' into v0.8 (+11 more commits...) - https://github.com/joyent/node/commit/a23df7710ab89a926d105e52a903bd044c7cea2d
19:27:19  * stagasquit (Ping timeout: 244 seconds)
19:27:31  * stagas_changed nick to stagas
19:27:35  <TooTallNate>oh shebang, y u no relative?
19:29:08  <creationix>too bad you can't do #!/bin/sh `pwd`/foo
19:29:30  <isaacs>TooTallNate: srsly.
19:29:31  <CIA-134>node: Bert Belder cares-move * r5b1e7b3 / (92 files in 12 dirs): Add c-ares dependency - https://github.com/joyent/node/commit/5b1e7b3171dd26d8c3458cc3e1306da3be9fc580
19:29:31  <CIA-134>node: Bert Belder cares-move * rf54f25e / (node.gyp src/cares_wrap.cc src/tree.h): Don't rely on libuv for c-ares integration - https://github.com/joyent/node/commit/f54f25e51c54972c88ff5b9a97ec2715d2d9bb48
19:29:42  <TooTallNate>creationix: you can do relative to pwd, but that's pretty much useless
19:29:43  <isaacs>piscisaureus: \o/
19:29:50  <isaacs>piscisaureus: are we gonna get js dns soon?
19:29:53  <creationix>TooTallNate, true, it's __dirname you want!
19:29:57  <TooTallNate>yup
19:30:01  <piscisaureus>isaacs: dunno. But this is a good first step :-)
19:30:09  <isaacs>agreed
19:30:25  <creationix>getting __dirname that works cross-sh is hard-enough, let-alone baked into a shebang line
19:30:58  <creationix>isaacs, I thought node had enough primitives for dns servers already?
19:31:18  <TooTallNate>creationix: hence https://github.com/joyent/node/commit/7a9db6cfb139a6cb5941b2ae09ab5671e474e160
19:31:25  <isaacs>TooTallNate: you have access: ssh [email protected]
19:31:38  <isaacs>TooTallNate: the dist folder is [email protected]:dist/
19:31:55  <isaacs>TooTallNate: standard warning: DONT BREAK THE WEBSITE!
19:32:12  <TooTallNate>isaacs: kewl thanks, i figure we could add a `make bin-upload` or something
19:32:26  <tjfontaine>creationix: so long as dgram doesn't go away :)
19:32:30  <TooTallNate>but we'll try it out so far as see how much work it takes
19:32:35  <isaacs>yeah, there's stll a few manual steps that i've been meaning to make-ify
19:32:46  <piscisaureus>creationix: cares should go away
19:32:55  <piscisaureus>creationix: you can keep it in luvit tho
19:33:06  <creationix>TooTallNate, oh, that's not too bad, but how portable is it? zsh/bash linux, osx
19:33:42  <TooTallNate>creationix: well `dirname "$0"` is pretty universal afaik
19:34:00  <creationix>I guess there are no symlinks involved in this case
19:34:05  <creationix>since npm is creating the files
19:34:28  <creationix>piscisaureus, I'm not attached to cares
19:34:42  <creationix>we use it because it's there, nothing more
19:34:49  <TooTallNate>well yay! binary packages are official now
19:34:54  <isaacs>sweet
19:34:58  * isaacsstarts v0.8.6-release
19:35:02  <TooTallNate>back to work-work :\
19:35:06  <creationix>binary packages?
19:35:33  <TooTallNate>creationix: we're gonna start releasing node-v0.8.6-linux-x86_64.tar.gz
19:35:38  <TooTallNate>and other common platforms
19:35:44  <creationix>rasp-pi!
19:35:48  <creationix>jk
19:36:06  <TooTallNate>i mean if there's a universal enough config for arm that will work almost everywhere then I'd be down
19:36:18  <TooTallNate>i haven't experimented enough with arm compiling yet though
19:36:26  <creationix>heh, portable arm binaries, I think that's an oxymoron
19:36:30  <TooTallNate>but i *do* have my pi now, so it's a possibility ;)
19:36:35  <TooTallNate>true
19:36:36  <TooTallNate>haha
19:37:32  <creationix>even just for a single hardware platform like rasp-pi where the kernel is pretty closed, there are variants of the os with and without hard-fp
19:40:24  <bnoordhuis>back
19:41:46  <CIA-134>node: isaacs v0.8 * r9badc00 / deps/uv/src/win/error.c : uv: Upgrade to dfb6be0 - https://github.com/joyent/node/commit/9badc002b4b109b237466c14de08278ef9be1e90
19:42:23  * `3rdEdenjoined
19:43:25  * stagasquit (Ping timeout: 250 seconds)
19:45:35  <piscisaureus>bnoordhuis: do you ming copying ares_inet_pton and ares_inet_ntop into the libuv source tree?
19:45:41  <piscisaureus>s/ming/mind/ ?
19:46:07  <bnoordhuis>piscisaureus: err, i guess i don't
19:46:21  <bnoordhuis>is it because you poor impoverished windows people don't have inet_pton and inet_ntop?
19:46:37  <piscisaureus>bnoordhuis: not on older windowses
19:46:55  <piscisaureus>bnoordhuis: i believe you poor unix souls have security issues with them btw
19:48:34  <bnoordhuis>piscisaureus: how so?
19:48:35  * c4miloquit (Read error: Connection reset by peer)
19:48:45  * c4milo_joined
19:49:18  * c4milo_quit (Read error: Connection reset by peer)
19:49:45  * c4milojoined
19:50:02  <piscisaureus>bnoordhuis: I don't really remember...
19:50:08  * c4miloquit (Read error: Connection reset by peer)
19:50:24  * c4milojoined
19:51:31  <piscisaureus>bnoordhuis: http://cxsecurity.com/issue/WLB-2008080064
19:51:38  <piscisaureus>bnoordhuis: maybe it's already dated
19:54:04  <bnoordhuis>piscisaureus: oh, that. old
19:55:54  * dylangjoined
20:05:05  * EhevuTovjoined
20:07:31  * EhevuTov_quit (Ping timeout: 240 seconds)
20:08:40  <CIA-134>node: isaacs v0.8 * rcaf69aa / (198 files in 9 dirs): npm: Upgrade to 1.1.47 - https://github.com/joyent/node/commit/caf69aa91a94c7c05574afc9fcc97f45aa8d5ea5
20:09:24  * indexzerojoined
20:11:55  <CIA-134>node: isaacs v0.8.6-release * r7bd6b6f / deps/uv/src/win/error.c : uv: Upgrade to dfb6be0 - https://github.com/joyent/node/commit/7bd6b6fe7eaacb5ef31d1f13d666bf65c8dedafa
20:12:01  <CIA-134>node: isaacs v0.8.6-release * r947f29f / (AUTHORS ChangeLog src/node_version.h): 2012.08.07, Version 0.8.5 (Stable) - https://github.com/joyent/node/commit/947f29f562ffee523102f164aa04f78b7deb841d
20:12:35  <mscdex>0.8.5?
20:12:47  <mscdex>:-)
20:13:15  <piscisaureus>isaacs: 0.8.6 !!
20:13:21  <mscdex>hehe
20:13:40  <piscisaureus>isaacs: s/UV_ENOE/UV_ENOENT/
20:14:12  <mscdex>bnoordhuis: so, it looks like the http parser error PR i sent is actually a bug *fix* since http parser errors were already supposed to be emitting on ClientRequests, it's just that not all of them were being emitted there
20:15:19  <mscdex>it's just that the http parser error happens for a different reason
20:16:10  * mcavagejoined
20:16:17  * mcavagequit (Remote host closed the connection)
20:16:24  <piscisaureus>ha! freebsd is on github
20:16:25  * mcavagejoined
20:16:26  <piscisaureus>yay
20:16:35  <mcavage>isaacs: hi :)
20:16:59  <isaacs>TooTallNate: https://gist.github.com/3278074
20:17:02  <isaacs>mcavage: ^
20:17:30  <isaacs>it looks like if os.environ['PORTABLE']:
20:17:31  <mcavage>isaacs: yep, same error as me
20:17:41  <isaacs>mcavage: same home directory as you even!! ;)
20:17:49  <TooTallNate>well my python skill are like 1
20:17:52  <TooTallNate>out of 10
20:17:53  <isaacs>TooTallNate: it looks like os.environ['PORTABLE'] blows up if it's not set
20:17:56  <isaacs>yeah..
20:17:56  <mcavage>lol
20:18:00  <isaacs>0.5 better than mine
20:18:00  <TooTallNate>haha
20:18:07  <mcavage>i didn't catch that you just stole mine ;)
20:18:51  <tjfontaine>if 'PORTABLE' in os.environ:
20:19:32  <tjfontaine>dunno if that satisfies it entirely
20:19:34  * Benviejoined
20:19:39  <TooTallNate>we need to use .get()
20:19:44  <mcavage>PORTABLE=1 make install works, fwiw
20:20:23  <TooTallNate>mcavage: from SO: "os.environ is a dictionary. Trying to access a key not present in the dictionary will throw a KeyError. The get method simply returns None when the key does not exists."
20:21:01  <mcavage>sweet!
20:22:06  <mcavage>that sounds bat shit crazy for behavior (i.e., foo['bar'] => throws, but foo.get('bar') doesn't). but, whatever. I have level 0 for python skills.
20:22:22  <tjfontaine>well, python is not javascript :)
20:22:30  <tjfontaine>it follows their data model
20:23:39  * AndreasMadsenquit (Remote host closed the connection)
20:23:40  <mcavage>i see. level 0. what do i know.
20:24:42  <TooTallNate>review bnoordhuis, isaacs, anyone? https://github.com/TooTallNate/node/commit/fix-make-install
20:24:43  <isaacs>https://github.com/isaacs/node/commit/14ba3627476c6092552ec69283e2aabc60e01afc.patch
20:24:48  <isaacs>oh,hahah
20:24:58  <isaacs>TooTallNate: i like yours better
20:25:05  <TooTallNate>:p
20:25:11  <TooTallNate>ok landing
20:25:30  <CIA-134>node: Nathan Rajlich v0.8 * r093be8b / tools/install.py : install: prevent a KeyError from being thrown when PORTABLE is not set - https://github.com/joyent/node/commit/093be8b16bb153d18bfec752d42f8b79a8958800
20:25:37  <tjfontaine>isaacs: oy :)
20:25:40  <isaacs>TooTallNate: works
20:25:46  <isaacs>thanks
20:25:57  <TooTallNate>no thank you :)
20:26:49  * dylangquit (Quit: dylang)
20:27:36  <mcavage>thanks guys.
20:28:33  <CIA-134>node: isaacs v0.8.6-release * r4317a70 / (AUTHORS ChangeLog src/node_version.h): 2012.08.07, Version 0.8.5 (Stable) - https://github.com/joyent/node/commit/4317a7086b148f26c1a92bafe5f237a59c069ba4
20:29:29  * dylangjoined
20:29:45  * loladirojoined
20:29:54  * theColejoined
20:31:26  <isaacs>rc0 uploading
20:31:49  * stagasjoined
20:32:49  * xaqquit (Read error: Connection reset by peer)
20:33:16  * xaqjoined
20:33:47  * bnoordhuisquit (Quit: Leaving)
20:36:28  <TooTallNate>isaacs: you didn't fix piscisaureus' comments
20:38:49  <bradleymeck>why does listen(port) not accept on ::1?
20:40:14  * mmalecki1away]quit (Ping timeout: 246 seconds)
20:44:33  <mscdex>bradleymeck: possibly because the default ip is '0.0.0.0' ?
20:46:19  <bradleymeck>mscdex: just odd that it doesn't grab the loopback, i would have thought it listened on ipv6 and 4
20:46:31  <mscdex>guess not
20:47:25  <TooTallNate>idk if node is fully ipv6 ready
20:47:31  <TooTallNate> / libuv
20:47:36  <isaacs>piscisaureus: typo, thanks
20:47:45  <isaacs>copy-paste error with narrow columns :)
20:48:16  <bradleymeck>TooTallNate: well i can listen manually on ::1 at least and it works, just doesnt like the default INADDR_ANY for IPv6 it looks like
20:49:23  * bnoordhuisjoined
20:50:55  <CIA-134>node: Josh Erickson master * r7b367a9 / (lib/net.js test/simple/test-net-isip.js): net: add support for IPv6 addresses ending in :: - https://github.com/joyent/node/commit/7b367a93ce126d4bf5a56e10ee44613c59ffbe84
20:52:34  <TooTallNate>bradleymeck: maybe that helps your problem ^
20:53:05  <TooTallNate>oh no, ::1 was already being tested for
20:53:51  <CIA-134>node: isaacs v0.8.6-release * r00ac2a3 / (AUTHORS ChangeLog src/node_version.h): 2012.08.07, Version 0.8.6 (Stable) - https://github.com/joyent/node/commit/00ac2a36b7820af5c0cd40f5731c6b1aaadd21de
20:54:55  <isaacs>uploaded RC1
20:55:18  <bradleymeck>::0 seems to work in all test cases i can see too ...
21:00:11  * EhevuTov_joined
21:03:01  * EhevuTovquit (Ping timeout: 245 seconds)
21:04:10  * dapquit (Quit: Leaving.)
21:08:07  * dapjoined
21:16:42  * bradleymeckquit (Quit: bradleymeck)
21:18:50  * xaqquit (Read error: Connection reset by peer)
21:19:21  * xaqjoined
21:22:26  * dshaw_quit (Quit: Leaving.)
21:31:47  * dylangquit (Remote host closed the connection)
21:32:07  * dylangjoined
21:38:54  <CIA-134>node: Ben Noordhuis master * r32e0027 / test/simple/test-net-socket-timeout.js : test: fix simple/test-net-socket-timeout - https://github.com/joyent/node/commit/32e00272c8ad55afebba5599445e982951d5ce44
21:42:38  <bnoordhuis>^ grr
21:42:49  <piscisaureus>bnoordhuis: these patches should be reverted
21:43:10  <piscisaureus>bnoordhuis: (the not listening until a listener is added)
21:43:22  <piscisaureus>bnoordhuis: this problem needs to be resolved differently
21:43:27  <bnoordhuis>agreed
21:43:38  <piscisaureus>bnoordhuis: isaacs does too (this was discussed yesterday)
21:43:49  <bnoordhuis>okay, i'll revert it then
21:43:56  <bnoordhuis>problem is it doesn't revert cleanly
21:44:12  <piscisaureus>sad
21:45:37  <bnoordhuis>guess i'll just revert all commits
21:49:23  * c4miloquit (Read error: Connection reset by peer)
21:49:34  * c4milojoined
21:51:16  * c4miloquit (Remote host closed the connection)
21:56:10  <bnoordhuis>okay, force-pushing in a minute - don't worry, it's only to eradicate the last commit
21:56:12  * dshaw_joined
21:56:31  <CIA-134>libuv: Bert Belder inet * ra700b7d / (LICENSE include/uv.h uv.gyp src/inet.c): Add libuv-ified versions of inet_pton and inet_ntop - https://github.com/joyent/libuv/commit/a700b7deaaf3e34d430f13c9affd83956e61fe53
21:56:32  <CIA-134>libuv: Bert Belder inet * r7c7595a / src/uv-common.c : Use uv_inet_ntop/uv_inet_pton, instead of the c-ares implementations - https://github.com/joyent/libuv/commit/7c7595a98a9bb5da6f6be87fb81530735fc9ad55
21:56:39  <piscisaureus>bnoordhuis ^-- wanna review that?
21:57:21  <bnoordhuis>sure, one sec
21:57:30  * travis-cijoined
21:57:30  <travis-ci>[travis-ci] joyent/libuv#525 (inet - 7c7595a : Bert Belder): The build failed.
21:57:30  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/a700b7deaaf3^...7c7595a98a9b
21:57:30  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2051912
21:57:30  * travis-cipart
21:57:35  <bnoordhuis>[01:43|% 100|+ 445|- 0]: Done <- woo!
21:57:49  <CIA-134>node: Ben Noordhuis master * r4c150ca / (7 files in 3 dirs): net: fix listen() regression, revert patches - https://github.com/joyent/node/commit/4c150ca0d0cc4bd8dc6d182e0bf850daf27a2422
21:58:53  * EhevuTov__joined
22:01:26  * `3rdEdenquit (Quit: Zzzzz)
22:02:13  <bnoordhuis>piscisaureus: i'd rather use the system pton/ntop functions on unices
22:02:18  * EhevuTov_quit (Ping timeout: 248 seconds)
22:02:43  <bnoordhuis>then again...
22:02:49  <stagas>how about require falling back to a fs.readFileSync utf8
22:03:27  <bnoordhuis>the reason i say that is that inet_pton() reportedly supports link-local addresses, which the c-ares version doesn't
22:06:45  <KiNgMaR>there could also be a version switch to use the native functions on Windows NT 6.x ;)
22:07:15  * theColequit (Quit: theCole)
22:08:30  <piscisaureus>KiNgMaR: yeah, but what benefit does it have?
22:09:21  <CIA-134>libuv: Ben Noordhuis inet * r40e6351 / src/inet.c : inet: fixups, squash - https://github.com/joyent/libuv/commit/40e6351b7a2469f371deb0263784e780e40467c5
22:10:16  <bnoordhuis>funny, i accidentally pushed a branch named origin/inet. github accepts it but it doesn't show up in the timeline
22:10:20  <isaacs>bnoordhuis: yes, piscisaureus speaks my mind here.
22:10:20  * travis-cijoined
22:10:21  <travis-ci>[travis-ci] joyent/libuv#526 (inet - 40e6351 : Ben Noordhuis): The build is still failing.
22:10:21  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/7c7595a98a9b...40e6351b7a24
22:10:21  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2052029
22:10:21  * travis-cipart
22:10:41  <piscisaureus>bnoordhuis: but that's okay - I can move inet.c to src/win
22:11:02  <piscisaureus>bnoordhuis: so you want simple wrappers in src/unix/inet.c?
22:11:26  <bnoordhuis>piscisaureus: well... let's go with your patches for now
22:11:31  <piscisaureus>ok
22:11:42  <piscisaureus>bnoordhuis: so it compiles on linux (w/ your fixups) ?
22:11:42  <bnoordhuis>it's easy enough to fix later on if the need arises
22:11:45  <bnoordhuis>yep
22:11:48  <piscisaureus>good
22:11:54  * rendarquit
22:12:34  <piscisaureus>bnoordhuis: ok, next stop is remove cares
22:12:35  <piscisaureus>doing it now
22:13:14  <CIA-134>libuv: Bert Belder inet * rd81992a / (src/inet.c src/uv-common.c): Use uv_inet_ntop/uv_inet_pton, instead of the c-ares implementations - https://github.com/joyent/libuv/commit/d81992a7fd64b5ef7cf02bc86d09d4652d1df12e
22:13:55  <bnoordhuis>piscisaureus: ^ commit log
22:14:14  <piscisaureus>huh, crap
22:15:47  <CIA-134>libuv: Bert Belder inet * r34893f0 / (LICENSE include/uv.h uv.gyp src/inet.c): Add libuv-ified versions of inet_pton and inet_ntop - https://github.com/joyent/libuv/commit/34893f076c3f56ae7f0589e6d5f101f0c383c24c
22:15:47  <CIA-134>libuv: Bert Belder inet * r75dd64a / src/uv-common.c : Use uv_inet_ntop/uv_inet_pton, instead of the c-ares implementations - https://github.com/joyent/libuv/commit/75dd64a448782de17cf6eade2967c62ef627aef4
22:19:41  * dylangquit (Ping timeout: 252 seconds)
22:20:26  * travis-cijoined
22:20:26  <travis-ci>[travis-ci] joyent/libuv#527 (inet - d81992a : Bert Belder): The build is still failing.
22:20:26  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/40e6351b7a24...d81992a7fd64
22:20:26  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2052056
22:20:26  * travis-cipart
22:28:39  * travis-cijoined
22:28:39  <travis-ci>[travis-ci] joyent/libuv#528 (inet - 75dd64a : Bert Belder): The build is still failing.
22:28:39  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/d81992a7fd64...75dd64a44878
22:28:39  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2052089
22:28:39  * travis-cipart
22:31:47  <CIA-134>libuv: Bert Belder inet * rc4d9319 / (5 files in 2 dirs): Remove c-ares tests and benchmarks - https://github.com/joyent/libuv/commit/c4d9319cd7cccee6cb06d668212a184d14076188
22:38:36  * EhevuTov__quit (Ping timeout: 265 seconds)
22:41:13  <piscisaureus>isaacs: could it be that v0.6.21 didn't set IS_RELEASE flag to 1 ?
22:41:55  <isaacs>piscisaureus: yes, that's exactly what happened
22:42:16  <isaacs>piscisaureus: there was notany other problem with it though
22:42:28  <piscisaureus>isaacs: right, ok
22:42:30  <isaacs>i'll release a 0.6.22 when the EDONTCARE flag clears.
22:42:40  <piscisaureus>someone is confused here by the -pre
22:42:47  <isaacs>yeh
22:42:54  <isaacs>tell them that '-pre' means that they should be using 0.8
22:42:55  <piscisaureus>so now I know what it is, I don't care either
22:42:56  <isaacs>:)
22:43:20  <isaacs>becasue clearly 0.6 is not our priority any more.
22:43:22  * EhevuTov__joined
22:45:14  <piscisaureus>oh, I thought...
22:45:23  <piscisaureus>0.6 is python 2
22:48:32  * travis-cijoined
22:48:32  <travis-ci>[travis-ci] joyent/libuv#529 (inet - c4d9319 : Bert Belder): The build is still failing.
22:48:32  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/75dd64a44878...c4d9319cd7cc
22:48:32  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2052236
22:48:32  * travis-cipart
22:58:49  * indexzeroquit (Quit: indexzero)
22:59:47  * indexzerojoined
23:00:02  * indexzeroquit (Client Quit)
23:06:39  <CIA-134>libuv: Bert Belder master * r3b46285 / config-mingw.mk : windows: remove libeio mentions from Makefile - https://github.com/joyent/libuv/commit/3b46285ff8dee216666b98e508b21392c351b75a
23:07:06  <CIA-134>libuv: Bert Belder inet * r32f93e1 / (6 files in 3 dirs): Add libuv-ified versions of inet_pton and inet_ntop - https://github.com/joyent/libuv/commit/32f93e14cbeea53ac8e2fc1dc77207dc6b251003
23:07:06  <CIA-134>libuv: Bert Belder inet * r3a8bb3b / src/uv-common.c : Use uv_inet_ntop/uv_inet_pton, instead of the c-ares implementations - https://github.com/joyent/libuv/commit/3a8bb3b2b1aae1d254eee0da62f55a2a8e921011
23:07:08  <CIA-134>libuv: Bert Belder inet * r35c4858 / (5 files in 2 dirs): Remove c-ares tests and benchmarks - https://github.com/joyent/libuv/commit/35c4858231b6f38d16e9c8a64cb0f14e9c1919b4
23:07:08  <CIA-134>libuv: Bert Belder inet * ra069956 / (8 files in 5 dirs): Remove c-ares integrations - https://github.com/joyent/libuv/commit/a06995699622fd25125373e1992b8ceffd85efb1
23:07:08  <CIA-134>libuv: Bert Belder inet * r41b1265 / (93 files in 11 dirs): Rip out c-ares - https://github.com/joyent/libuv/commit/41b1265af8329131154539cb0d1eda57758b62be
23:07:17  <piscisaureus>^-- bnoordhuis isaacs: any last words?
23:07:40  <bnoordhuis>goodnight sweet prince, we hardly knew ye?
23:09:39  * travis-cijoined
23:09:40  <travis-ci>[travis-ci] joyent/libuv#530 (master - 3b46285 : Bert Belder): The build is still failing.
23:09:40  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/32da23988bd4...3b46285ff8de
23:09:40  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2052486
23:09:40  * travis-cipart
23:10:10  * travis-cijoined
23:10:10  <travis-ci>[travis-ci] joyent/libuv#531 (inet - 41b1265 : Bert Belder): The build was fixed.
23:10:10  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/c4d9319cd7cc...41b1265af832
23:10:10  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2052488
23:10:10  * travis-cipart
23:11:26  <bnoordhuis>piscisaureus: lgtm
23:11:36  <piscisaureus>it even builds on linux :-)
23:11:41  <KiNgMaR>piscisaureus: re: native InetPtonW: if there ever was - god forbid - a security issue in libuv's own pton implementation, Windows Vista and later might get off lucky :)
23:11:52  <CIA-134>libuv: Bert Belder master * r32f93e1 / (6 files in 3 dirs): Add libuv-ified versions of inet_pton and inet_ntop - https://github.com/joyent/libuv/commit/32f93e14cbeea53ac8e2fc1dc77207dc6b251003
23:11:52  <CIA-134>libuv: Bert Belder master * r3a8bb3b / src/uv-common.c : Use uv_inet_ntop/uv_inet_pton, instead of the c-ares implementations - https://github.com/joyent/libuv/commit/3a8bb3b2b1aae1d254eee0da62f55a2a8e921011
23:11:53  <CIA-134>libuv: Bert Belder master * r35c4858 / (5 files in 2 dirs): Remove c-ares tests and benchmarks - https://github.com/joyent/libuv/commit/35c4858231b6f38d16e9c8a64cb0f14e9c1919b4
23:11:53  <CIA-134>libuv: Bert Belder master * ra069956 / (8 files in 5 dirs): Remove c-ares integrations - https://github.com/joyent/libuv/commit/a06995699622fd25125373e1992b8ceffd85efb1
23:11:54  <CIA-134>libuv: Bert Belder master * r41b1265 / (93 files in 11 dirs): Rip out c-ares - https://github.com/joyent/libuv/commit/41b1265af8329131154539cb0d1eda57758b62be
23:12:28  * bradleymeckjoined
23:14:22  * travis-cijoined
23:14:22  <travis-ci>[travis-ci] joyent/libuv#532 (master - 41b1265 : Bert Belder): The build was fixed.
23:14:22  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/3b46285ff8de...41b1265af832
23:14:22  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/2052523
23:14:22  * travis-cipart
23:26:49  * xaqquit (Remote host closed the connection)
23:42:07  * loladiroquit (Remote host closed the connection)
23:46:49  <bnoordhuis>TooTallNate: the commit header should be <= 60 chars
23:46:59  <bnoordhuis>i.e. the first line of the commit log
23:47:23  * bradleymeckquit (Quit: bradleymeck)
23:51:06  <CIA-134>node: Bert Belder cares-move * r27af1ff / (102 files in 15 dirs): uv: upgrade to 41b1265 (+5 more commits...) - https://github.com/joyent/node/commit/27af1ffc4dcb12991dff00f41ca39681d972eb2d
23:53:05  <CIA-134>node: Bert Belder cares-move * r735897d / (node.gyp src/cares_wrap.cc src/tree.h): dns: don't rely on libuv for c-ares integration - https://github.com/joyent/node/commit/735897d900353d8f5ca625f94994bd5f0f692512
23:53:06  <CIA-134>node: Bert Belder cares-move * r9c54576 / (28 files in 7 dirs): uv: upgrade to 3a8bb3b - https://github.com/joyent/node/commit/9c54576c91e33323c269915622d8decf76b36e1f
23:53:06  <CIA-134>node: Bert Belder cares-move * rf9c0e91 / (src/cares_wrap.cc src/tcp_wrap.cc src/udp_wrap.cc): dns: use uv_inet_ntop/uv_inet_pton instead of c-ares equivalents - https://github.com/joyent/node/commit/f9c0e91c49d8cb0060230a663a6ebe2fc87814f6
23:53:06  <CIA-134>node: Bert Belder cares-move * reb9e87e / (102 files in 15 dirs): uv: upgrade to 41b1265 - https://github.com/joyent/node/commit/eb9e87e7c3024b0fb87e398aa32faf3a722227a1