00:06:07  * SouL_quit (Remote host closed the connection)
00:22:00  * DarkGodquit (Ping timeout: 250 seconds)
00:31:17  * pfraze_joined
00:53:41  * hdmsjoined
01:09:12  * erlbot--quit (Remote host closed the connection)
01:09:28  * erlbot--joined
03:07:49  * pfraze_quit (Remote host closed the connection)
03:09:48  * pfrazejoined
05:17:57  * jetlquit (Read error: Connection reset by peer)
05:31:05  * imzyxwvujoined
05:37:17  * imzyxwvuquit (Ping timeout: 246 seconds)
06:00:06  * hdmsquit (Quit: hdms)
06:31:26  * pfrazequit (Remote host closed the connection)
06:56:52  * kinbajoined
06:58:00  * kinbaquit (Remote host closed the connection)
06:58:52  * kinjoined
07:01:31  * kinchanged nick to kinba
07:22:39  * kinbaquit
07:38:28  * DarkGodjoined
07:59:58  * hdmsjoined
08:00:21  * hdmsquit (Client Quit)
10:20:37  * rphillipsquit (Quit: ZNC - http://znc.in)
10:21:07  * rphillipsjoined
10:22:31  * Akagi201_quit
10:47:13  * DarkGodquit (Ping timeout: 255 seconds)
10:48:47  * joconnorjoined
11:02:16  * torporjoined
11:02:38  * DarkGodjoined
11:28:07  * imzyxwvujoined
11:32:26  * SouLjoined
12:59:00  * imzyxwvuquit (Ping timeout: 244 seconds)
14:11:30  * pfrazejoined
14:31:48  <rphillips>good morning
14:45:02  * pfrazequit (Remote host closed the connection)
15:07:12  * pfrazejoined
15:21:19  * travis-cijoined
15:21:20  <travis-ci>luvit/luv#298 (master - 9115e39 : Ryan Phillips): The build has errored.
15:21:20  <travis-ci>Change view : https://github.com/luvit/luv/compare/b5e0e43de3b4...9115e394e407
15:21:20  <travis-ci>Build details : http://travis-ci.org/luvit/luv/builds/56897266
15:21:20  * travis-cipart
15:29:25  * travis-cijoined
15:29:26  <travis-ci>luvit/luv#298 (master - 9115e39 : Ryan Phillips): The build passed.
15:29:26  <travis-ci>Change view : https://github.com/luvit/luv/compare/b5e0e43de3b4...9115e394e407
15:29:26  <travis-ci>Build details : http://travis-ci.org/luvit/luv/builds/56897266
15:29:26  * travis-cipart
15:29:46  <creationix>rphillips: do we use the zlib bindings for anything? We never implemented the node sugar on top and the zip asset code in luvi used miniz
15:30:06  <creationix>I’m considering dropping the zlib option from the prebuilt binaries
15:30:07  <rphillips>not that I know of
15:30:11  <rphillips>+1
15:30:45  <creationix>that should help shrink the binary a little (though nothing like openssl)
15:34:13  <rphillips>7.5 MB RSS for agent2 running overnight
15:34:30  <rphillips>not bad at all
15:47:01  * hdmsjoined
15:56:43  <rphillips>http://xkcd.com/1506/
16:20:37  <creationix>rphillips: so luvi release today?
16:20:41  <rphillips>+`1
16:20:42  <rphillips>1
16:21:34  <creationix>shall I update lua-openssl to get your mode_bitmask pr?
16:22:26  <creationix>hmm, looks like you already did, nevermind
16:39:37  * travis-cijoined
16:39:38  <travis-ci>luvit/luvi#471 (master - 815a053 : Tim Caswell): The build was broken.
16:39:38  <travis-ci>Change view : https://github.com/luvit/luvi/compare/550ab82c07a2...815a053fe5ee
16:39:38  <travis-ci>Build details : http://travis-ci.org/luvit/luvi/builds/56908996
16:39:38  * travis-cipart
16:46:00  * travis-cijoined
16:46:01  <travis-ci>luvit/luvi#472 (master - 784305a : Tim Caswell): The build passed.
16:46:01  <travis-ci>Change view : https://github.com/luvit/luvi/compare/815a053fe5ee...784305abb27f
16:46:01  <travis-ci>Build details : http://travis-ci.org/luvit/luvi/builds/56909333
16:46:01  * travis-cipart
16:47:16  * travis-cijoined
16:47:17  <travis-ci>luvit/luvi#473 (master - 6cc1fd4 : Tim Caswell): The build has errored.
16:47:17  <travis-ci>Change view : https://github.com/luvit/luvi/compare/784305abb27f...6cc1fd43a16e
16:47:17  <travis-ci>Build details : http://travis-ci.org/luvit/luvi/builds/56910124
16:47:17  * travis-cipart
16:48:23  <rje>sorry i missed you guys yesterday
16:48:32  <rje>rphillips, re: distro 14.04
17:06:12  * DarkGodquit (Ping timeout: 256 seconds)
17:12:40  * travis-cijoined
17:12:41  <travis-ci>luvit/luvi#475 (v1.1.0 - bf603f5 : Tim Caswell): The build passed.
17:12:41  <travis-ci>Change view : https://github.com/luvit/luvi/compare/v1.1.0
17:12:41  <travis-ci>Build details : http://travis-ci.org/luvit/luvi/builds/56913924
17:12:41  * travis-cipart
17:19:37  * joconnorquit (Remote host closed the connection)
17:25:58  <creationix>New luvi release. All builds except for linux_arm6l are complete
17:26:18  <creationix>the arm6l is taking extra long since my builtbot went up in smoke yesterday and I need to setup another on my older pi
17:26:21  * pfrazequit (Remote host closed the connection)
17:26:35  <creationix>https://github.com/luvit/luvi/releases/tag/v1.1.0
17:27:08  <creationix>The Raspberry PI 2 is so much faster, it completed before I could start all the desktop builds
17:27:19  * pfrazejoined
17:30:17  <rje>was it the one with the supply issues?
17:30:56  <creationix>they’ve all had supply issues when they were first announced
17:31:08  <creationix>but the pi2 wasn’t too bad, they were generally available after a month
17:31:31  <creationix>the original pi took me 6 months to get and that was after staying up all night and hitting refresh toe 8 hours straight to get in on the first order
17:32:02  * pfrazequit (Ping timeout: 250 seconds)
17:47:27  <rje>can we merge this yet? https://github.com/luvit/luvit/pull/669
17:48:27  <creationix>rje: what does flushheaders do?
17:48:50  <creationix>is that for http trailers or something?
17:51:08  <rje>just like node.js the headers aren't written until either post data is sent or the request is marked done. flushheaders writes the headers, its called by the first write or request done
17:52:10  <creationix>shouldn’t it be called unconditionally then?
17:52:32  <creationix>also I think the closing “” should be written all the time too. Did you say it breaks if you try to write an empty string in non-chunked mode?
17:54:07  <rje>it is called unconditionally in the case of the Write()
17:54:26  <creationix>right, but write may be called zero times
17:54:41  <creationix>what if I do res.setHeader() res.finish(“some data”)?
17:55:11  <rje>https://github.com/luvit/luvit/blob/http_post_error/deps/http.lua#L376
17:55:43  <rje>that'll flush the headers there
17:55:58  <creationix>oh, that write
17:56:21  <rje>?
17:56:29  <creationix>hmm, but you call _done later on with the last data chunk? https://github.com/luvit/luvit/blob/http_post_error/deps/http.lua#L385
17:56:37  <creationix>you can’t send body after writing empty to http-codec
17:56:57  <rphillips>rje: i updated the waffle tracker
17:56:59  <rphillips>with a bunch of stuff
17:57:30  <rje>rphillips: that's what i'm reading through. thanks for doing that
17:57:46  <rphillips>i got the ntp stuff within the agent fixed
17:57:57  <rphillips>buildbots are pushing to cloudfiles
17:59:40  <rje>creationix: i'm thinking about that... that done may be mis-ordered
18:00:16  <creationix>my understanding of :finish is
18:00:17  <creationix>1 - flush headers
18:00:18  <creationix>2 - write data if there is some
18:00:19  <creationix>3 - write empty string to codec to signify end of stream
18:00:32  <creationix>I’m afraid I don’t understand much else, sorry if I’m confusing things
18:00:57  <creationix>unconditionally writing an empty string to the codec should work for EOS, if not, it’s a bug that needs fixing
18:02:08  <rje>yup, what i need to do there is to write the data passed in done first https://github.com/luvit/luvit/blob/http_post_error/deps/http.lua#L376
18:04:21  * torpor1joined
18:05:01  <rje>then write the empty string into the _done
18:05:14  <creationix>that sounds good
18:05:44  * DarkGodjoined
18:06:09  * torporquit (Ping timeout: 252 seconds)
18:07:02  * torpor1quit (Client Quit)
18:07:09  * torporjoined
18:07:09  * torporquit (Client Quit)
18:07:13  * torpor1joined
18:14:38  * travis-cijoined
18:14:39  <travis-ci>luvit/luvi#476 (master - d17348e : Tim Caswell): The build passed.
18:14:39  <travis-ci>Change view : https://github.com/luvit/luvi/compare/bf603f5acb88...d17348e61485
18:14:39  <travis-ci>Build details : http://travis-ci.org/luvit/luvi/builds/56923210
18:14:39  * travis-cipart
18:31:57  * torpor1quit (Quit: Leaving.)
18:34:12  * pfrazejoined
18:41:30  * joconnorjoined
18:42:12  * pfrazequit (Remote host closed the connection)
18:54:36  * SouLquit (Read error: Connection reset by peer)
19:00:02  <rje>creationix: https://github.com/luvit/luvit/commit/8381f5fd32ee0fb8725f50c850da64e8d91bb665
19:03:32  * travis-cijoined
19:03:33  <travis-ci>luvit/luvit#1917 (http_post_error - 8381f5f : Rob Emanuele): The build was broken.
19:03:34  <travis-ci>Change view : https://github.com/luvit/luvit/compare/0e642e376c51...8381f5fd32ee
19:03:34  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/56930642
19:03:34  * travis-cipart
19:08:22  <creationix>rje: I wonder if that https-client error is related
19:08:31  <creationix>the patch does look much better though
19:09:41  <rje>that's what would fail when i always send '' to the codec
19:10:10  <creationix>interesting
19:11:06  <rje>can we get "Missing 1 expected call" to print the line numbers of the missing calls?
19:11:10  <rphillips>https://github.com/virgo-agent-toolkit/rackspace-monitoring-agent/pull/682
19:11:15  <rphillips>i'm going to merge this... sound ok?
19:11:52  <creationix>rje: I’m not sure how to get that kind of debug info
19:11:55  <rje>rphillips: if you don't want me to finish moving it to base......
19:12:00  <creationix>but I can debug the test if you want
19:12:19  <rphillips>rje: it's ok... i'll need to tweak it a tad for the command line argument
19:12:34  <rje>creationix: if i could get another set of eye's on that i'd be great
19:13:07  <rje>rphillips: then merge it
19:13:12  <rphillips>coo
19:21:41  <creationix>rje: ok, the problem is that sometimes the encoder will emit empty strings and we need to not write that to the socket
19:23:41  <creationix>rje: this helps a lot https://github.com/luvit/luvit/commit/1d205d79c474719605118a253dbd18cb642cff56
19:24:18  <rje>scweet
19:24:46  * joconnorquit (Remote host closed the connection)
19:26:16  * travis-cijoined
19:26:18  <travis-ci>luvit/luvit#1919 (http_post_error - 1d205d7 : =): The build was fixed.
19:26:18  <travis-ci>Change view : https://github.com/luvit/luvit/compare/8381f5fd32ee...1d205d79c474
19:26:18  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/56934735
19:26:18  * travis-cipart
19:27:05  * torporjoined
19:27:28  * torporquit (Client Quit)
19:32:15  <creationix>rje: I added one more tweak to not allow empty chunks going to the encoder when they are meant as data. That would cause confusion in chunked encoding mode
19:36:25  <rje>creationix: great, there's one other issue with the case where the done is called before connected
19:40:14  <rje>creationix: i feel like this version of data should be encoded https://github.com/luvit/luvit/pull/669/files#diff-ca5fe4df66831f999037cad7605bf824L373
19:40:38  <creationix>I have no idea what self.ended is for
19:40:42  <creationix>but yes, probably
19:42:29  <rje>it handles teminating the connection if the connection hasn't been set up yet, https://github.com/luvit/luvit/blob/http_post_error/deps/http.lua#L334
19:42:39  <creationix>hmm, I wonder if :_done can handle empty strings properly because the encoder will emit those
19:43:13  <rje>hmm, with that change [string "/Users/rje/raxvc/luvit/deps/http-codec.lua"]:94: expected a table but got a string when encoding data
19:43:44  <creationix>so what I do know:
19:43:45  <creationix>1 - passing “” to the codec means EOS, you should do this exactly once for every request
19:43:45  <creationix>2 - sometimes the encoder emits empty strings, you should always ignore these
19:45:39  * travis-cijoined
19:45:40  <travis-ci>luvit/luvit#1921 (http_post_error - 52851c5 : =): The build has errored.
19:45:40  <travis-ci>Change view : https://github.com/luvit/luvit/compare/1d205d79c474...52851c58fe99
19:45:40  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/56935864
19:45:40  * travis-cipart
19:47:43  <creationix>technically the encoder only needs the empty string for EOS when using chunked encoding, but since the decision to use chunked encoding is internal to the codec, it’s best to always send it instead of trying to duplicate the detection logic
19:58:23  * pfrazejoined
20:27:36  <creationix>rphillips: https://github.com/luvit/lit/pull/57
20:46:52  <creationix>hmm, doesn’t update the luvi portion
21:26:56  * erlbot--quit (Remote host closed the connection)
21:27:04  * erlbot--joined
21:27:16  <creationix>alright, now lit make will looks for luvi version and flavor in package.lua and downloads that version if the current executable doesn’t match
21:27:16  <creationix>so lit updates will get luvi updates
21:27:20  <creationix>also we can lit make using tiny flavor
21:27:25  <creationix>for apps that don’t need openssl
21:27:58  <rphillips>that is a cool feature
21:34:13  <rphillips>lockfile is working... tweaked it slightly to not load the ffi on windows
21:38:47  * creationixquit (Ping timeout: 265 seconds)
21:39:07  <rphillips>an example binding for weblit using the websocket module would be slick
21:40:35  * creationixjoined
22:49:24  * pfrazequit (Remote host closed the connection)
22:51:53  * pfrazejoined
22:53:42  * pfrazequit (Remote host closed the connection)
23:16:31  * travis-cijoined
23:16:32  <travis-ci>luvit/luvit#1923 (http_post_error - fbb98f3 : Rob Emanuele): The build passed.
23:16:32  <travis-ci>Change view : https://github.com/luvit/luvit/compare/52851c58fe99...fbb98f321576
23:16:32  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/56969217
23:16:32  * travis-cipart
23:18:21  * travis-cijoined
23:18:22  <travis-ci>luvit/luvit#1925 (http_post_error - 7957fdc : Rob Emanuele): The build passed.
23:18:22  <travis-ci>Change view : https://github.com/luvit/luvit/compare/fbb98f321576...7957fdcb631d
23:18:22  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/56969476
23:18:22  * travis-cipart
23:44:21  <rje>creationix: this looks better, i had to rollback one of your changes though, https://github.com/luvit/luvit/commit/fbb98f32157643d7d7795e452960ecce401a1c2c
23:44:41  <rje>for https://github.com/luvit/luvit/pull/669
23:51:18  * pfrazejoined