00:14:12  <philips>I just tried running an ab benchmark on the simple http server of nodejs vs luvit and node was giving me 6000 r/s and luvit was 5100 r/s
00:14:14  <philips>0.6.18 nodejs
00:14:16  <philips>master luvit
00:14:37  <philips>memory usage was like 50 MB peak vs 3.2 MB peak
00:14:41  <philips>so we got that I guess
00:18:53  <creationix>philips: node was 50mb?
00:22:04  <philips>creationix: yea, it kept peaking around there
00:22:11  <philips>creationix: have you not seen that?
00:22:21  <creationix>I believe it
00:22:24  <creationix>node is a memory hog
00:22:31  <creationix>just making sure that wasn't luvit
00:22:41  <philips>creationix: oh, heh
00:23:00  <creationix>I wonder why luvit was so slow though
00:23:04  <creationix>was that a 32bit vm?
00:24:20  <philips>creationix: no, it was 64bit
00:24:30  <creationix>that's a regression
00:24:31  <philips>Linux shell.ifup.org 2.6.32-5-xen-amd64 #1 SMP Mon Jan 16 20:48:30 UTC 2012 x86_64 GNU/Linux
00:24:38  <philips>creationix: :-/
01:05:41  * tsingjoined
01:33:58  * TheJHquit (Read error: Operation timed out)
02:26:03  <rphillips>did the string refs go in?
03:40:38  * hij1nxjoined
06:04:34  * hij1nxquit (Quit: hij1nx)
06:05:01  * hij1nxjoined
06:19:30  * hij1nxquit (Quit: hij1nx)
06:44:08  * dvvjoined
06:54:35  * dvvquit (Ping timeout: 244 seconds)
07:03:12  * mmaleckijoined
07:38:02  * mmaleckiquit (Read error: Connection reset by peer)
08:11:25  * mmaleckijoined
08:26:57  * mmaleckiquit (Read error: Connection reset by peer)
09:19:45  * mmaleckijoined
10:00:27  * `3rdEdenjoined
10:12:21  * `3rdEdenquit (Quit: Leaving...)
10:44:19  * tsingquit (Quit: Linkinus - http://linkinus.com)
10:59:47  * mmaleckiquit (Ping timeout: 240 seconds)
11:10:10  * mmaleckijoined
11:16:31  * TheJHjoined
11:41:33  * mmaleckiquit (Quit: leaving)
12:13:44  * tsingjoined
13:02:25  <creationix>string refs?
13:39:03  * 18VAAM8FVjoined
13:47:24  * ljacksonquit (*.net *.split)
13:48:23  * ljacksonjoined
14:39:51  * hij1nxjoined
14:48:08  <rphillips>philips mentioned adding string refs so the GC doesn't reap them when we are in an async call
15:16:38  * tsingquit (*.net *.split)
15:16:58  * tsingjoined
15:25:46  * tsingquit (Quit: Linkinus - http://linkinus.com)
15:36:13  * AvianFlujoined
15:37:13  * hij1nxquit (Quit: hij1nx)
15:40:44  * hij1nxjoined
16:41:33  * hij1nxquit (Quit: hij1nx)
17:08:34  * hij1nxjoined
17:28:46  * `3rdEdenjoined
17:28:49  * tsingjoined
17:43:47  * tsingquit (Quit: Linkinus - http://linkinus.com)
17:57:20  <philips>Slides from my 5 minute talk last night: http://ifup.org/slides/luvit-5mof/#1
17:57:37  <rphillips>how did it go?
17:57:40  <rphillips>interest?
17:57:50  <philips>rphillips: I had fun, I think people were interested
17:57:54  <rphillips>awesome
17:58:05  <philips>rphillips: We should probably put analytics on luvit.io
17:58:14  <rphillips>+1
17:59:31  <rphillips>would be nice to figure out the performance hit
17:59:38  <philips>rphillips: yea...
17:59:47  <philips>rphillips: I knew we used to be better but I didn't want to lie
18:02:04  <rphillips>philips: hmm. the tables, tables slide was to show OO w/ tables?
18:02:25  <philips>rphillips: yea, and like functions are table members, etc
18:02:36  <rphillips>would be cool to show the OO module for luvit
18:02:39  <philips>rphillips: Just that the tables data structure is used everywhere for everything
18:02:49  <rphillips>gotcha
18:03:03  <philips>rphillips: Yea, I was mostly explaining the language at that point- didn't want to get into how there is no standard OO system in lua...
18:04:01  <rphillips>i see
18:04:06  <rphillips>i love the octocat slide
19:04:52  * `3rdEdenquit (Quit: Leaving...)
19:31:40  * `3rdEdenjoined
19:33:04  * `3rdEdenquit (Client Quit)
20:07:10  <pquerna>rphillips: hrm, i don't know what you mean by luvit alrady unrefs it
20:07:30  <pquerna>rphillips: if i do a setInterval() in luvit, it'll exit the event loop if thats the only one left?
20:09:14  <rphillips>pquerna: hmm
20:09:43  <rphillips>you are right
20:09:50  <rphillips>need a week interval timer
20:10:12  <rphillips>a regular timer unrefs itself
20:10:20  <pquerna>oh?
20:10:24  <pquerna>even that seems weird right/
20:10:37  <rphillips>not really... it's what node does
20:10:50  <rphillips>it unrefs the timer and only refs the timer when it's active
20:11:03  <pquerna>hmmm
20:11:18  <pquerna>so a hello world print inside a setTimeout won't work?
20:11:22  <pquerna>it'd work in node.js afaik?
20:11:46  <rphillips>it works in luvit also
20:12:07  <pquerna>confusedddd then
20:12:44  <rphillips>i'm looking it up... it was confusing when I ported it from node
20:14:26  <rphillips>pquerna: https://github.com/luvit/luvit/blob/master/lib/luvit/uv.lua#L188
20:14:36  <rphillips>there is a native.timerGetActive call
20:14:57  <rphillips>which figures out if the timer is active in the event loop... depending on the state of the timer it will ref or unref it
20:15:19  <rphillips>so by default the timer handle gets unref'ed
20:16:23  <rphillips>https://github.com/joyent/node/blob/master/src/timer_wrap.cc#L93
20:18:21  <pquerna>ah, except when active, i get it i think
20:18:27  <pquerna>uv_timer_init shouldn't do it.. right
20:18:51  <rphillips>uv_timer_init refs it... you're right, it's questionable if it should do it
20:19:02  <pquerna>okay
20:19:18  <pquerna>so we still don't have 'weak' timers to do 'internal' things like forcing GC
20:21:29  <rphillips>pquerna: actually, I propose we call uv.unref()
20:21:45  <pquerna>rphillips: yeah, we can just do that in userspace i guess
20:21:51  <pquerna>for this one use case
20:21:52  <rphillips>it'llhttps://gist.github.com/2af4a0934341623e7fe3
20:21:53  <pquerna>seems better?
20:21:54  <rphillips>https://gist.github.com/2af4a0934341623e7fe3
20:21:57  <rphillips>it works
20:22:20  <pquerna>did we want to move the GC timer into Luvit mainline?
20:22:31  <pquerna>but this looks good, the gist
20:22:32  <pquerna>+1
20:23:05  <rphillips>preference thing moving it to mainline.
20:23:33  <rphillips>there should be an api to make it tunable
20:32:30  <rphillips>pquerna: +1 on the PR?
20:47:03  <pquerna>looking
20:48:29  <pquerna>rphillips: +1
20:48:37  <rphillips>sweet. thx
21:00:52  * 18VAAM8FVquit (Quit: Leaving.)
21:10:02  * AvianFluquit (Quit: Leaving)