00:09:18  * a__quit (Remote host closed the connection)
00:11:48  * a__joined
00:47:07  * a__quit (Remote host closed the connection)
00:49:20  * a__joined
00:56:23  * a__quit (Remote host closed the connection)
01:03:59  * a__joined
01:04:59  * kazuponjoined
01:12:05  * a__quit (Remote host closed the connection)
01:12:28  * DarkGodquit (Quit: Leaving)
01:26:16  * kazuponquit (Remote host closed the connection)
01:38:12  * dan336joined
01:44:22  * kazuponjoined
02:09:30  * a__joined
02:14:00  * a__quit (Remote host closed the connection)
02:17:15  * a__joined
02:25:20  * a__quit (Remote host closed the connection)
02:33:02  * a__joined
02:36:52  * a__quit (Remote host closed the connection)
02:38:01  * a__joined
02:48:20  * a__quit (Remote host closed the connection)
03:25:44  * kazuponquit (Remote host closed the connection)
04:02:47  * kazuponjoined
04:12:24  * dan336quit (Quit: Leaving.)
07:09:48  * kazuponquit (Remote host closed the connection)
07:21:06  * kazuponjoined
08:13:19  * torporjoined
08:15:25  * srenatusjoined
09:00:49  * torpor1joined
09:03:14  * torporquit (Ping timeout: 256 seconds)
09:15:50  * ra_quit (Ping timeout: 244 seconds)
09:18:37  * ra_joined
10:11:35  * kazuponquit (Remote host closed the connection)
12:45:26  * UniOnjoined
13:13:27  * kazuponjoined
13:21:18  * kazuponquit (Remote host closed the connection)
13:55:30  * kazuponjoined
15:16:02  <rphillips>hmm. still getting a handle is already closing error
15:16:45  <creationix>rphillips: where?
15:16:54  <rphillips># Starting Test: socket-timeout - socket timeout
15:16:58  <rphillips>on osx
15:17:14  <creationix>yeah, I’ve seen some strangeness on that test too
15:17:19  <rphillips>the socket is getting cancelled
15:17:20  <rphillips>ECANCELED: operation canceled
15:27:53  <creationix>rphillips: how do you reproduce the problem. It runs pretty well on my mac
15:28:21  <rphillips>just make test...
15:28:28  <rphillips>i think i found the issue
15:29:23  <creationix>I did notice that on windows, it only passes the first time I run the test. Every other run has issues
15:32:47  <rphillips>creationix: can you try this branch on windows? fixes/fix_ecancelled
15:33:04  <rphillips>fixed it on osx for me
15:33:47  <rphillips>minor tweaks to the tap module as well
15:33:57  <rphillips>https://github.com/luvit/luvit/pull/524/files
15:34:43  * travis-cijoined
15:34:43  <travis-ci>luvit/luvit#1064 (http-codec - b2e68a7 : Tim Caswell): The build passed.
15:34:43  <travis-ci>Change view : https://github.com/luvit/luvit/compare/030ab41d5a1c...b2e68a708aac
15:34:43  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40552460
15:34:43  * travis-cipart
15:35:26  * travis-cijoined
15:35:26  <travis-ci>luvit/luvit#1065 (tls-codec - fa75452 : Tim Caswell): The build is still failing.
15:35:26  <travis-ci>Change view : https://github.com/luvit/luvit/compare/577aa718c1cc...fa7545218a06
15:35:26  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40552703
15:35:26  * travis-cipart
15:36:05  * travis-cijoined
15:36:06  <travis-ci>luvit/luvit#1066 (fixes/fix_ecancelled - 1cd4f95 : Ryan Phillips): The build passed.
15:36:06  <travis-ci>Change view : https://github.com/luvit/luvit/commit/1cd4f9511402
15:36:06  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40552948
15:36:06  * travis-cipart
15:36:31  <rphillips>i wish appveyor was a bit faster on starting a build
15:40:24  * dan336joined
15:46:00  <rphillips>still queued... 13 minutes :(
15:49:40  * a_lequit (Remote host closed the connection)
15:50:00  * kazuponquit (Read error: Connection reset by peer)
15:50:27  * kazuponjoined
16:00:14  <creationix>yep, this is why I keep an old windows laptop on my desk
16:06:37  * a_lejoined
16:09:40  * kazuponquit (Remote host closed the connection)
16:12:06  * dan3361joined
16:12:07  * dan336quit (Read error: Connection reset by peer)
16:21:36  <creationix>rphillips: would it be improper for luv to check is_closing inside uv.close?
16:21:43  <creationix>luv is meant to be “barebones"
16:22:31  <rphillips>no, I think that would be a good fix
16:27:44  <creationix>well, the problem is the callback would then be called sync or not at all. Either is strange behavior for uv.close
16:28:32  <rphillips>yeah, and after further thought the uv bindings really should be raw bindings I think
16:50:05  * jirwinquit (Changing host)
16:50:06  * jirwinjoined
17:16:48  <rphillips>creationix: want me to tweak that PR?
17:29:26  * a_lequit (Remote host closed the connection)
17:59:45  <rphillips>tweaked
17:59:49  <rphillips>didn't see your comment before
17:59:53  <rphillips>removed the tap changes
18:05:19  * travis-cijoined
18:05:19  <travis-ci>luvit/luvit#1068 (fixes/fix_ecancelled - c5dcd76 : Ryan Phillips): The build has errored.
18:05:19  <travis-ci>Change view : https://github.com/luvit/luvit/compare/1cd4f9511402...c5dcd760cceb
18:05:19  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40568828
18:05:19  * travis-cipart
18:22:39  <rphillips>much better
18:22:50  <rphillips>didn't happen everytime on osx either
18:30:00  * travis-cijoined
18:30:00  <travis-ci>luvit/luvit#1071 (luvi-up - 21c303d : Ryan Phillips): The build passed.
18:30:00  <travis-ci>Change view : https://github.com/luvit/luvit/compare/5be8070670de...21c303d9a088
18:30:00  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40571548
18:30:00  * travis-cipart
18:33:09  * srenatusquit (Quit: Connection closed for inactivity)
18:49:20  * a_lejoined
18:50:12  * dan3361quit (Quit: Leaving.)
18:51:47  * a_lequit (Read error: Connection reset by peer)
18:52:08  * a_lejoined
18:54:50  * travis-cijoined
18:54:50  <travis-ci>luvit/luvit#1072 (add-emitter - 6f7a3e2 : Ryan Phillips): The build passed.
18:54:50  <travis-ci>Change view : https://github.com/luvit/luvit/commit/6f7a3e23780a
18:54:50  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40574100
18:54:50  * travis-cipart
19:01:43  * travis-cijoined
19:01:44  <travis-ci>luvit/luvit#1074 (add-emitter - e1d85b2 : Ryan Phillips): The build passed.
19:01:44  <travis-ci>Change view : https://github.com/luvit/luvit/compare/6f7a3e23780a...e1d85b270425
19:01:44  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40574740
19:01:44  * travis-cipart
19:12:30  <creationix>rphillips: so what do you think about the style of the http-codec code?
19:12:51  <creationix>It seems to perform fine and I'd prefer not include a C binding if we don't need it. Lua code is easier to work with.
19:14:04  <rphillips>my only concern is that most projects use http-parser or curl, and we have custom code
19:14:10  <rphillips>but I think it's a blocker
19:14:13  <rphillips>don't think*
19:14:35  <creationix>worried about bugs in the new code?
19:14:42  <rphillips>right
19:15:10  <creationix>I've been using the js version in js-git without issues and I'm willing to fix bugs as we find them
19:15:44  <creationix>I don't want to reimplement tls in lua. As much as I'd love such a library to exist, I'm not going to attempt it myself
19:16:27  <creationix>my issues with http_parser and curl is they are slow to implement new protocols like spdy or http2
19:17:29  <rphillips>plus we will want a native http parser
19:17:43  <rphillips>i'm game... I do like the interface
19:17:59  <rphillips>should be easy to write other protocols
19:18:59  <creationix>Did you see the difference between my tranducer style approach (emitter -> emitter transform) and the coroutine-style (blocking read, blocking write)
19:19:18  <creationix>the latest code is the latter style. It's clearer I think and supports backpressure much better
19:20:20  <rphillips>link?
19:24:00  <creationix>rphillips: here is the new style https://github.com/luvit/luvit/blob/http-codec/bench/http-cluster/app.lua#L1
19:24:17  <creationix>read blocks till there is data to read, and write blocks if the receiver isn't ready to take it
19:24:20  <creationix>so instant backpressure
19:25:49  <creationix>the old style takes an emitter (a callback that is called to push events) and returns a new emitter https://github.com/luvit/luvit/blob/4bd100ffce91864b97582cc69d8f77b2166a6a37/test-http-server.lua#L7-L8
19:26:17  <creationix>backpressure has to be handled externally and you have to worry about being re-entrant and keeping state between events
19:26:34  <creationix>https://github.com/luvit/luvit/blob/4bd100ffce91864b97582cc69d8f77b2166a6a37/test-http-server.lua#L7-L8
19:26:47  * creationixthinks some of my messages got dropped by the network
19:28:44  <rch>it all made sense to me
19:32:36  <rphillips>same here
19:45:18  * a_lequit (Remote host closed the connection)
19:45:58  * a_lejoined
19:48:37  * a_lequit (Read error: Connection reset by peer)
19:48:55  * a_lejoined
19:52:19  * a_lequit (Remote host closed the connection)
19:52:54  * a_lejoined
20:02:02  * a_lequit (Read error: Connection reset by peer)
20:02:23  * a_lejoined
20:07:35  * torpor1quit (Quit: Leaving.)
20:10:39  * dan336joined
20:25:34  <rphillips>https://github.com/luvit/luvit/pull/526
20:25:41  <rphillips>got the crypto tests ported to the new openssl module
20:35:36  <creationix>rphillips: awesome
20:35:46  <creationix>I commented on the readSync code.
20:35:51  * travis-cijoined
20:35:51  <travis-ci>luvit/luvit#1080 (port-crypto - ad89afe : Ryan Phillips): The build passed.
20:35:51  <travis-ci>Change view : https://github.com/luvit/luvit/compare/e6d578f5be3d...ad89afe0d3b1
20:35:51  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40585048
20:35:51  * travis-cipart
20:35:51  <creationix>Sorry I forgot to add that
20:36:46  <rphillips>np. I'll add error handling
20:59:10  <rphillips>ok. added
21:00:15  <rphillips>hmm
21:00:18  <rphillips>that isn't quite right
21:04:22  <rphillips>creationix: ok... i think that looks better
21:05:24  <creationix>rphillips: or, sorry I didn't explain well. luv doesn't throw for libuv errors, it uses assert-style return values
21:05:44  <creationix>I think the only times I throw are for things like you passed in a string when the bindings expected a number
21:05:50  <creationix>(before it gets to libuv)
21:06:09  <creationix>local fd, err = fs.openSync(...)
21:11:12  * travis-cijoined
21:11:12  <travis-ci>luvit/luvit#1084 (port-crypto - 9da8aa8 : Ryan Phillips): The build passed.
21:11:12  <travis-ci>Change view : https://github.com/luvit/luvit/compare/0f621270c09b...9da8aa805f96
21:11:12  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40588553
21:11:12  * travis-cipart
21:11:25  * travis-cijoined
21:11:25  <travis-ci>luvit/luvit#1082 (port-crypto - 0f62127 : Ryan Phillips): The build has errored.
21:11:25  <travis-ci>Change view : https://github.com/luvit/luvit/compare/ad89afe0d3b1...0f621270c09b
21:11:25  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40587966
21:11:25  * travis-cipart
21:13:20  <rphillips>creationix: doesn't it throw luaL_errors?
21:14:36  <creationix>not for uv stuff
21:15:10  <creationix>it returns nil, err
21:15:20  <creationix>but if you wrap it in assert, then it throws
21:15:39  <rphillips>creationix: can you show me an example of an error path?
21:16:29  <rphillips>or, perhaps what that function show be?
21:25:32  <creationix>rphillips: https://github.com/luvit/luvit/pull/526/files#r20113943
21:25:38  <rphillips>oh, the 2nd param is the error
21:25:48  <rphillips>got it
21:25:50  <rphillips>thanks
21:31:01  <rphillips>updated
21:31:18  <rphillips>i'm digging the new openssl bindings
21:31:45  <creationix>cool
21:39:16  <creationix>rphillips: btw, this pr from Saturday is ready for a review https://github.com/luvit/luvit/pull/523
21:44:02  <rphillips>creationix: commented on two minor nits... need license headers on some of the files
21:44:21  <rphillips>epic PR
21:45:22  <creationix>thanks
21:57:50  * travis-cijoined
21:57:50  <travis-ci>luvit/luvit#1088 (http-codec - 83f299f : Tim Caswell): The build passed.
21:57:50  <travis-ci>Change view : https://github.com/luvit/luvit/compare/b2e68a708aac...83f299f76f31
21:57:50  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40592849
21:57:50  * travis-cipart
22:09:07  <rphillips>still 1 second for unit tests
22:20:02  * typedlambdaquit (Ping timeout: 250 seconds)
22:22:15  * typedlambdajoined
22:22:20  <creationix>rphillips: how’s this for docs on the cluster example: https://github.com/luvit/luvit/blob/43a23fc9c2d539f2e2efa0133e8c05053e648167/bench/http-cluster/README.md
22:24:08  <rphillips>damn. nice.. that is more than I thought
22:24:09  <rphillips>+1
22:28:43  <creationix>rphillips: I’m thinking of changing the structure of the req/res tables to be flat
22:29:09  <creationix>the numerical indexes of the headers don’t conflict with the named properties like code, method, version, etc
22:29:23  <creationix>basically just inline the headers list in the main table
22:29:35  <creationix>lua makes this fairly painless I think
22:30:06  * travis-cijoined
22:30:07  <travis-ci>luvit/luvit#1092 (http-codec - ea7de89 : Tim Caswell): The build passed.
22:30:07  <travis-ci>Change view : https://github.com/luvit/luvit/compare/43a23fc9c2d5...ea7de894dd87
22:30:07  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40596171
22:30:07  * travis-cipart
22:31:05  * travis-cijoined
22:31:06  <travis-ci>luvit/luvit#1090 (http-codec - 43a23fc : Tim Caswell): The build has errored.
22:31:06  <travis-ci>Change view : https://github.com/luvit/luvit/compare/83f299f76f31...43a23fc9c2d5
22:31:06  <travis-ci>Build details : http://travis-ci.org/luvit/luvit/builds/40595562
22:31:06  * travis-cipart