00:52:24  * Akagi201joined
01:18:02  * DarkGodquit (Ping timeout: 244 seconds)
02:29:03  * dan336joined
04:27:35  * dan336quit (Quit: Leaving.)
04:35:24  * SkyRocknRollquit (Ping timeout: 244 seconds)
05:19:05  * SkyRocknRolljoined
06:13:05  * sousouxquit (Ping timeout: 246 seconds)
06:13:48  * sousouxjoined
07:44:06  * DarkGodjoined
08:08:05  * sousouxquit (Ping timeout: 256 seconds)
08:49:51  * DarkGodquit (Remote host closed the connection)
08:51:20  * DarkGodjoined
08:54:45  * DarkGodquit (Remote host closed the connection)
08:56:51  * DarkGodjoined
10:10:58  * torporjoined
10:31:35  * Akagi201quit (Remote host closed the connection)
13:17:05  * Akagi201joined
13:17:40  * SkyRocknRollquit (Ping timeout: 264 seconds)
13:18:14  * Akagi201_joined
13:22:01  * Akagi201quit (Ping timeout: 264 seconds)
13:23:52  <rphillips>good morning
13:33:28  * SkyRocknRolljoined
13:46:14  * torporquit (Quit: Leaving.)
14:03:01  * sousouxjoined
14:07:48  * SkyRocknRollquit (Ping timeout: 272 seconds)
14:21:31  * SkyRocknRolljoined
14:30:16  * dan336joined
14:40:36  * piernov__joined
14:43:48  * squeeekjoined
14:46:36  * piernovquit (*.net *.split)
14:46:41  * squeekquit (*.net *.split)
15:33:23  <creationix>good morning
15:36:03  <rphillips>things are looking good
16:22:35  * SkyRocknRollquit (Remote host closed the connection)
16:56:51  * torporjoined
16:56:56  * torporquit (Client Quit)
17:33:35  * sousouxquit (Ping timeout: 248 seconds)
17:54:36  * sousouxjoined
19:13:32  <sousoux>rphillips: saw your message. unfortunately just calling shutdown on an ssl socket and then destroying is not enough. SSL has a shutdown handshake (to avoid MITM DoS attack). If you shutdown like that then the session is invalidated at the browser end. If you do it repeatedly you get into problems. I have a patch for this but I still have another issue with libuv closing the (or at least transmitting
19:13:32  <sousoux>an EOF) socket just after the handshake on my device (not on PC Linux) which I'm still tracking down. I'll put in the patch when I find that one. For now tracking that one down is priority #1.
19:14:20  <sousoux>rphillips: you have to repeatedly call shutdown and pump data to and from the BIO until shutdown tells you all is ok.
19:14:49  * sousouxquit
19:14:49  <rphillips>gotcha. ok
19:15:08  * sousouxjoined
19:16:13  <sousoux>And I don't see _destroyed anywhere else so that is not doing much
19:16:42  <rphillips>it's in the net.lua file as well
19:22:13  <sousoux>ok but the point above is absolutely valid
19:22:18  <sousoux>https://www.openssl.org/docs/ssl/SSL_shutdown.html
19:22:20  <rphillips>yeah
19:22:26  <rphillips>a patch for that would be great
19:22:41  <rphillips>i didn't realize you had to pump the ssl_shutdown
19:22:56  <sousoux>On PC linux I could not get repeated web requests to a server to work until I fixed it
19:24:03  <sousoux>Now onto the libuv issue which is a PITA. The browser really does not like TLS sockets being repeatedly reset.
19:26:32  <sousoux>Its reaction is to invalidate the session each time and in the end give up
19:27:47  <sousoux>I have an (unconfirmed) feeling that it may be related to the stop/start just after the handshake completes
20:38:48  <sousoux>rphillips: in net.lua the only action on _destroyed is to set it false on initialize. Everywhere else destroyed (without the underscore) is used.
20:39:25  <rphillips>hmm. that sounds like a bug
21:04:52  * bjornquit (Quit: No Ping reply in 180 seconds.)
21:05:51  * b_lindeijerjoined
21:05:51  * b_lindeijerquit (Changing host)
21:05:51  * b_lindeijerjoined
22:46:57  * dan336quit (Quit: Leaving.)
23:02:07  * sousouxquit (Ping timeout: 248 seconds)
23:02:23  * sousouxjoined
23:07:06  * Akagi201_quit (Remote host closed the connection)
23:17:44  * erlbot--quit (Remote host closed the connection)
23:41:48  * erlbot--joined