00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:07  * ircretaryjoined
00:06:33  * indexzerojoined
00:09:21  * indexzeroquit (Client Quit)
00:09:38  * mikealjoined
00:21:49  * indexzerojoined
00:23:35  * ralquit (Ping timeout: 240 seconds)
00:26:21  * icarot`joined
00:26:56  * icarot`quit (Remote host closed the connection)
00:27:19  * bnoordhuisquit (Ping timeout: 260 seconds)
00:29:56  * raljoined
00:29:56  * ralquit (Changing host)
00:29:56  * raljoined
00:30:24  * icarotquit (Remote host closed the connection)
00:31:46  <tjfontaine>ok https://github.com/joyent/node/pull/5500 is ready for review
00:31:53  * Benviequit
00:32:02  <tjfontaine>and I really need to fix my chrome pull request plugin
00:34:49  * TooTallNatequit (Quit: Computer has gone to sleep.)
00:41:43  * indexzeroquit (Quit: indexzero)
00:42:16  * mikealquit (Quit: Leaving.)
00:42:31  * cjdquit (Ping timeout: 264 seconds)
00:54:46  * TooTallNatejoined
00:55:52  * indexzerojoined
00:59:28  * inolenquit (Quit: Leaving.)
01:03:40  * dapquit (Quit: Leaving.)
01:03:50  * indexzeroquit (Quit: indexzero)
01:06:29  * amartensjoined
01:07:37  * TooTallNatequit (Ping timeout: 276 seconds)
01:08:52  * TooTallNatejoined
01:17:46  * abraxasjoined
01:20:12  * TooTallNatequit (Quit: Computer has gone to sleep.)
01:24:13  * brsonquit (Quit: leaving)
01:33:04  * bnoordhuisjoined
01:34:30  * inolenjoined
01:37:54  * bnoordhuisquit (Ping timeout: 264 seconds)
02:09:36  * dannycoatesquit (Remote host closed the connection)
02:20:16  * Benviejoined
02:22:18  * amartensquit (Quit: Leaving.)
02:37:59  * dshaw_joined
02:44:50  * brsonjoined
02:51:49  * mikealjoined
02:54:50  * wmiljoined
03:01:01  * indexzerojoined
03:14:14  * benoitcquit (Excess Flood)
03:18:32  * benoitcjoined
03:27:51  * wmilquit (Remote host closed the connection)
03:37:22  * indexzeroquit (Quit: indexzero)
03:39:08  * amartensjoined
03:40:16  * wmiljoined
03:51:05  * wmilquit (Remote host closed the connection)
03:54:38  * kazuponjoined
03:56:50  * defunctzombie_zzchanged nick to defunctzombie
04:02:43  * indexzerojoined
04:20:00  * normanmjoined
04:26:43  * c4miloquit (Remote host closed the connection)
04:41:49  * amartensquit (Quit: Leaving.)
04:57:13  * wmiljoined
05:00:01  * wmilquit (Remote host closed the connection)
05:03:45  * kazupon_joined
05:03:57  * kazuponquit (Read error: Connection reset by peer)
05:33:13  * kazupon_quit (Remote host closed the connection)
05:35:13  * defunctzombiechanged nick to defunctzombie_zz
05:35:49  * paddybyersjoined
05:37:23  * kazuponjoined
06:05:18  * mikealquit (Quit: Leaving.)
06:12:00  * bajtosjoined
06:13:10  * mikealjoined
06:16:06  * mikealquit (Client Quit)
06:17:50  * brsonquit (Ping timeout: 256 seconds)
06:21:14  * `3rdEdenjoined
06:30:46  * mikealjoined
06:35:33  * rendarjoined
07:04:01  * stagasjoined
07:20:14  * csaohjoined
07:26:58  * icarotjoined
07:29:26  * csaohquit (Quit: csaoh)
07:30:45  * dominictarrjoined
07:33:12  * stagasquit (Ping timeout: 256 seconds)
07:41:27  * stagasjoined
07:46:32  * csaohjoined
07:49:16  * bajtosquit (Quit: bajtos)
08:16:44  * bajtosjoined
08:31:15  * timoxleyjoined
08:39:43  * bnoordhuisjoined
08:49:07  * wmiljoined
09:04:28  * stagasquit (Read error: Connection reset by peer)
09:20:59  <bajtos>bnoordhuis: ping (see private message)
09:22:08  * hzjoined
09:26:24  * dshaw_quit (Quit: Leaving.)
09:51:24  * wmilquit (Remote host closed the connection)
09:53:41  * indexzeroquit (Quit: indexzero)
09:56:45  * mraleph1joined
09:56:52  * mralephquit (Ping timeout: 252 seconds)
10:01:33  * wmiljoined
10:06:34  * timoxleyquit (Quit: Computer has gone to sleep.)
10:12:45  * wmilquit (Remote host closed the connection)
10:14:35  * timoxleyjoined
10:19:09  * abraxasquit (Remote host closed the connection)
10:28:45  * csaohquit (Quit: csaoh)
10:33:37  * timoxleyquit (Quit: Computer has gone to sleep.)
10:49:34  * hz_joined
10:49:35  * hz_quit (Changing host)
10:49:35  * hz_joined
10:49:35  * hzquit (Disconnected by services)
10:49:39  * hz_changed nick to hz
11:01:25  * csaohjoined
11:01:38  * kazuponquit (Remote host closed the connection)
11:02:25  * icarotquit (Ping timeout: 248 seconds)
11:41:39  * zz_jcequit (Excess Flood)
11:41:39  * benoitcquit (Excess Flood)
11:41:52  * zz_jcejoined
11:42:26  * zz_jcequit (Excess Flood)
11:42:52  * zz_jcejoined
11:43:52  * zz_jcequit (Excess Flood)
11:44:22  * zz_jcejoined
11:44:38  * zz_jcequit (Excess Flood)
11:44:52  * zz_jcejoined
11:44:56  * zz_jcequit (Excess Flood)
11:45:23  * zz_jcejoined
11:49:04  * benoitcjoined
12:07:54  * bnoordhuisquit (Ping timeout: 264 seconds)
12:12:06  * kazuponjoined
12:17:09  * kazuponquit (Ping timeout: 248 seconds)
12:19:25  * abraxasjoined
12:22:56  * bnoordhuisjoined
12:23:37  * abraxasquit (Ping timeout: 240 seconds)
12:34:49  * bajtosquit (Quit: bajtos)
12:43:45  * c4milojoined
12:46:00  * kazuponjoined
12:49:30  * kevinswiberjoined
12:51:11  * c4miloquit (Remote host closed the connection)
13:01:59  <indutny>hoya
13:10:28  * defunctzombie_zzchanged nick to defunctzombie
13:14:54  <bnoordhuis>sup fedor?
13:19:55  * defunctzombiechanged nick to defunctzombie_zz
13:23:23  * st_lukejoined
13:25:33  * kazuponquit (Remote host closed the connection)
13:34:04  * c4milojoined
13:36:02  * pachetjoined
13:36:02  * pachetquit (Changing host)
13:36:02  * pachetjoined
13:55:36  * st_lukequit (Remote host closed the connection)
14:17:18  * Benvie_joined
14:19:44  * abraxasjoined
14:20:47  * Benviequit (Ping timeout: 260 seconds)
14:24:19  * abraxasquit (Ping timeout: 252 seconds)
14:25:05  * cjdjoined
14:27:08  <cjd>hi guys, I'm using uv_pipe_t and I see node seems able to create anonymous bidirectional pipes but the only way I can do it is by creating AF_UNIX sockets which crap up /tmp every time my code crashes...
14:27:19  <cjd>Can someone point me in the direction of the anonymous ipc pipes?
14:27:33  <cjd>the js->c++ link is a little hard to follow
14:36:02  * kazuponjoined
14:37:29  <bnoordhuis>cjd: you mean socketpair(2)?
14:37:50  <bnoordhuis>you can pass one end of the socketpair to uv_pipe_open()
14:39:14  <cjd>oh and that's platform independent? perfect.
14:40:21  * kazuponquit (Ping timeout: 240 seconds)
14:41:07  * kazuponjoined
14:41:22  <tjfontaine>did I agree to run the call today?
14:41:39  <bnoordhuis>cjd: no, unix only
14:43:05  <bnoordhuis>tjfontaine: yes. yes, you did
14:43:15  <cjd>hmm, how does node create stdin and stdout pipes on windows?
14:43:19  <cjd>for child_process
14:43:50  <bnoordhuis>cjd: through platform logic in src/win/process.c
14:43:52  <tjfontaine>heh alright, I'll be in the office in about an hour
14:44:20  <bnoordhuis>cjd: uv-unix does similar things in src/unix/process.c (using, among other things, socketpairs)
14:45:11  <cjd>ok, I'll keep poking around
14:45:15  * piscisaureus_joined
14:49:16  * bajtosjoined
14:49:29  * mikealquit (Quit: Leaving.)
14:53:36  * st_lukejoined
14:55:56  * bajtosquit (Ping timeout: 252 seconds)
14:56:09  * dominictarrquit (Quit: dominictarr)
15:30:43  * `3rdEdenquit (Remote host closed the connection)
15:36:00  * TooTallNatejoined
15:44:02  * st_lukequit (Remote host closed the connection)
15:52:43  * cjdquit (Ping timeout: 264 seconds)
15:55:24  <tjfontaine>ok so
15:56:13  <isaacs>i'm doing this webinar thing this morning
15:56:22  * Benviejoined
15:56:33  <isaacs>indutny, bnoordhuis, piscisaureus_, sblom, trevnorris: tjfontaine will be leading the skype checkin
15:56:42  * dapjoined
15:56:47  <indutny>sorry, I'll skip today's call
15:56:48  <piscisaureus_>ok
15:56:55  * isaacshas been working on http keepalive exploration
15:57:10  <isaacs>and reviewing #4964
15:59:23  <bnoordhuis>there is this bearded man sending me invites on skype. i'm scared
15:59:48  <tjfontaine>what can I say
16:00:02  <tjfontaine>shall we start the call then?
16:00:21  <bnoordhuis>only if you promise to keep all your clothes on this time
16:02:21  * benoitcquit (Read error: Operation timed out)
16:08:05  * benoitcjoined
16:09:20  <bnoordhuis>trevnorris: i'm fine with the out-of-bounds patch
16:09:46  * mikealjoined
16:09:57  * amartensjoined
16:16:38  <trevnorris>bnoordhuis: thanks
16:18:39  * brsonjoined
16:20:23  * dannycoatesjoined
16:23:23  <piscisaureus_>Man I need to catch up on trev's buffer stuff
16:23:25  <bnoordhuis>and a lot of other things
16:23:25  <piscisaureus_>like what?
16:23:25  <bnoordhuis>bert still thinks v0.8.20 is the latest release
16:23:25  <tjfontaine>so does voxer for that matter :)
16:23:25  * kazuponquit (Remote host closed the connection)
16:25:14  * mikealquit (Quit: Leaving.)
16:26:14  <piscisaureus_>bnoordhuis: so you are talking to the android guy somewhere outside of github?
16:26:43  <tjfontaine>bnoordhuis: one thing I was thinking about while working on that timer stuff, how would you feel about only having two timers, one for ref'd and one for unref'd?
16:27:10  <bnoordhuis>piscisaureus_: yeah, over irc. it's kind of like voxer but without a/v
16:27:26  <bnoordhuis>tjfontaine: the fewer timers, the better
16:28:14  <tjfontaine>the management of the list scares me some, since we're not finding a bucket to place it in, but finding where in the list it should be placed, unless we went with a minheap or similar
16:28:23  * stagasjoined
16:33:11  <MI6>joyent/node: Trevor Norris v0.10 * 2cad7a6 : buffer: throw when writing beyond buffer - http://git.io/a_DZQw
16:34:40  * st_lukejoined
16:36:35  * bnoordhuisquit (Ping timeout: 260 seconds)
16:43:35  <MI6>nodejs-v0.10: #201 FAILURE linux-ia32 (1/584) http://jenkins.nodejs.org/job/nodejs-v0.10/201/
16:43:47  * stagasquit (Read error: Connection reset by peer)
16:44:12  <tjfontaine>hm that looks like github issues
16:47:08  <TooTallNate>ya definintely
16:47:11  <TooTallNate>they're being slow right now
16:47:28  <TooTallNate>https://github.com/joyent/node/pull/5515
16:47:36  <TooTallNate>^ in case anyone's looking for something to review :p
16:53:33  <trevnorris>sure. i'm waiting for the xbox reveal announcement, so there goes the morning. ;-)
16:55:06  * inolenquit (Quit: Leaving.)
16:55:36  * inolenjoined
16:55:46  <pachet>spoiler: its a hypercube
16:55:51  <trevnorris>lol
16:56:05  <trevnorris>TooTallNate: lgtm, though i'm not a python dev :P
16:56:28  * inolenquit (Client Quit)
16:58:16  <TooTallNate>trevnorris: ya me either :D
16:58:20  <TooTallNate>trevnorris: thanks though
17:01:02  <tjfontaine>depending on how this meeting goes I may review it during :)
17:02:57  * bnoordhuisjoined
17:05:13  <TooTallNate>trevnorris: you guys do lots of meetings, huh? :p
17:05:19  <TooTallNate>sorry, tjfontaine ^
17:05:23  <TooTallNate>damn autocomplete‚Ķ
17:06:05  * qardjoined
17:07:07  <tjfontaine>TooTallNate: tuesday is my meeting day, but otherwise it's just a daily standup
17:07:21  <TooTallNate>i see
17:09:32  * mikealjoined
17:12:13  * bnoordhuisquit (Ping timeout: 276 seconds)
17:14:02  * st_lukequit (Read error: Connection reset by peer)
17:16:54  * benoitcquit (Excess Flood)
17:17:20  <trevnorris>oy, daily standup. node standup is my only meeting.
17:17:43  * dominictarrjoined
17:19:35  * wmiljoined
17:22:07  * stagasjoined
17:24:04  * benoitcjoined
17:25:32  * wmilquit (Remote host closed the connection)
17:33:48  * kazuponjoined
17:34:29  * csaohquit (Quit: csaoh)
17:39:01  * kazuponquit (Ping timeout: 252 seconds)
17:44:02  * dshaw_joined
17:49:40  * inolenjoined
17:56:25  * pachetquit (Ping timeout: 276 seconds)
18:01:39  * benoitcquit (Excess Flood)
18:04:35  * brsonquit (Ping timeout: 252 seconds)
18:05:30  * brsonjoined
18:05:32  * brsonquit (Client Quit)
18:05:35  * benoitcjoined
18:05:45  * brsonjoined
18:07:17  * pachetjoined
18:07:36  * pachetquit (Client Quit)
18:11:21  * piscisaureus_quit (Read error: Operation timed out)
18:16:55  * mikealquit (Ping timeout: 260 seconds)
18:18:26  * sblomjoined
18:20:50  <sblom>Sorry I was AWOL for the call. It slipped my mind.
18:21:09  <sblom>tjfontaine: Any particularly firey Windows fires at present?
18:21:37  <tjfontaine>no huge fires I'm aware of at the moment, just the ones that have been simmering :)
18:23:27  <sblom>tjfontaine: Fair enough. I hope to find a bucket or two of water shortly.
18:24:06  <tjfontaine>sounds good, let me know if there's anything I can do to help
18:26:38  * `3rdEdenjoined
18:57:11  * brsonquit (Ping timeout: 256 seconds)
18:58:19  * brsonjoined
19:07:32  * bnoordhuisjoined
19:07:47  * brsonquit (Ping timeout: 260 seconds)
19:11:44  * bnoordhuisquit (Ping timeout: 245 seconds)
19:21:41  * rendarquit (Ping timeout: 248 seconds)
19:23:15  * toothrotquit (Ping timeout: 256 seconds)
19:23:18  * rendarjoined
19:23:56  <trevnorris>tjfontaine: i'm adding you to the wiki's project organization list. want to be listed as TJ?
19:26:54  * toothrjoined
19:33:41  * normanmquit (Quit: Computer has gone to sleep.)
19:39:44  <tjfontaine>trevnorris: ya that'll be fine
19:41:44  * st_lukejoined
19:49:53  <trevnorris>API and feedback. Buffer.prototype.<some_method> that will return a new Buffer instance which is not pooled (e.g. lives on its own)
19:50:35  <trevnorris>was going to use .yank(). I hate naming new api's
19:52:35  <tjfontaine>so it copies into a new buffer that's outside the pool?
19:52:42  <trevnorris>yeah
19:53:04  <tjfontaine>why can't they use .copy on their new buffer that came from outside the pool?
19:53:39  <trevnorris>because of the buffer is small then the new Buffer will also be part of a pool.
19:54:10  <trevnorris>basically since SlowBuffer's no longer exist there's no simply way for devs to create a small Buffer that's not in a pool.
19:54:55  <trevnorris>and people have complained about memory leaks from persisting pools, so this'll give them a way to control their own memory.
19:55:22  <trevnorris>e.g. stick everything in a pool for super speed, then the items you know will stick around can be yank'd out into their own
19:57:29  <tjfontaine>right, to me it seems like instead of adding a convenience method, you want a way to allocate a buffer outside the pool in general, then they can do the explicit copy on their own and others can put a convenience method in if they like it
19:58:20  <trevnorris>two things. there's no way to simply call "new Buffer" and have it return a small un-pooled buffer
19:58:39  <trevnorris>and it's useful if they've received some data from the SlabAllocator and want to quickly yank it out.
19:58:45  <tjfontaine>we couldn't have new Buffer(len[, escapeThePool])?
19:58:59  <trevnorris>no, second handles the encoding for strings.
19:59:12  <tjfontaine>we couldn't have new Buffer(len[, encoding[, escapeThePool]])?
19:59:21  <tjfontaine>:)
20:00:00  <trevnorris>tjfontaine: https://github.com/jmatthewsr-ms/node-slab-memory-issues
20:00:24  <tjfontaine>I'm familiar
20:00:53  <trevnorris>one of the points is to allow devs a way around this issue easily
20:01:09  <tjfontaine>I would rather approach it from the standpoint about moving away from our slab allocator and using libumem (or similar) instead of malloc and free
20:01:38  <trevnorris>that stuff only deals w/ maybe 5% of the performance impact of creating a Buffer.
20:01:58  <trevnorris>it's the persisting. making weak, gc object transversal that kills performance
20:02:19  <tjfontaine>that's not the point necessarily of it though
20:02:41  <trevnorris>the point of the slab allocator has nothing to do with libumem
20:03:05  <tjfontaine>I'm not communicating this well enough
20:03:24  <trevnorris>the slab allocator doesn't exist to reduce the number of malloc's. it's there to reduce the number of persistent js objects created.
20:07:29  * bnoordhuisjoined
20:09:36  <tjfontaine>bnoordhuis: #5500 if/when you get a chance
20:10:16  * st_lukequit (Read error: Connection reset by peer)
20:10:24  * st_lukejoined
20:11:33  <trevnorris>tjfontaine: so how could libumem/et all help with the performance problem that forces us to use pools?
20:11:56  <trevnorris>bbiab
20:11:57  * trevnorris&
20:11:57  <LOUDBOT>F00LI5H GET DOWN FROM THERE YOU ARE A CAT
20:15:11  <tjfontaine>trevnorris: I'm not saying it would help with that portion of the problem, just that the path that seems interesting to me to try and solve it is to remove the pools and to block allocate creation of the persistents, then let memory allocs happen through another mechanism perhaps with libumem
20:17:02  * AvianFluquit (Remote host closed the connection)
20:20:37  <txdv>libumem?
20:22:06  <txdv>an existing solaris lib
20:22:16  <txdv>i thought you want to create a libuv related lib that manages ... memory
20:22:27  <tjfontaine>txdv: https://github.com/gburd/libumem
20:22:56  * st_lukequit (Remote host closed the connection)
20:28:15  * brsonjoined
20:32:12  <MI6>nodejs-v0.10: #202 UNSTABLE smartos-x64 (2/584) http://jenkins.nodejs.org/job/nodejs-v0.10/202/
20:33:00  * bnoordhu1sjoined
20:34:57  * `3rdEdenquit (Quit: Zzzzz)
20:37:54  * bnoordhu1squit (Ping timeout: 264 seconds)
20:52:20  * wmiljoined
21:00:20  * mikealjoined
21:06:33  * rendarquit
21:19:32  * mikealquit (Quit: Leaving.)
21:22:07  * st_lukejoined
21:25:41  * kevinswiberquit (Remote host closed the connection)
21:26:17  * kevinswiberjoined
21:30:45  * kevinswiberquit (Ping timeout: 256 seconds)
21:30:59  * bnoordhuisquit (Remote host closed the connection)
21:31:31  <trevnorris>tjfontaine: well, i'm down for exploring it. :)
21:31:49  * bnoordhuisjoined
21:31:54  <tjfontaine>heh
21:33:10  * c4miloquit (Remote host closed the connection)
21:33:41  <trevnorris>oh and
21:33:42  * trevnorrisfg
21:36:01  * bnoordhuisquit (Ping timeout: 252 seconds)
21:44:42  * stagasquit (Read error: Connection reset by peer)
21:56:37  <trevnorris>anyone mind running the following:
21:56:42  <trevnorris>(on master)
21:56:54  <trevnorris>/usr/bin/time ./node benchmark/tls/throughput.js dur=10 type=buf size=32768
21:57:04  * piscisaureus_joined
21:57:04  <trevnorris>and showing the maxresident?
21:57:50  <trevnorris>I think tls has a pretty serious memory leak there
21:57:52  <trevnorris>indutny: ^
22:07:24  * st_lukequit (Remote host closed the connection)
22:36:31  <amartens>trevnorris: sure, just let me do a fresh checkout
22:47:46  <amartens>tls/throughput.js dur=10 type=buf size=32768: 381.59
22:47:46  <amartens> 10.16 real 8.88 user 1.25 sys
22:50:12  <trevnorris>amartens: hm looks like it didn't show your maxresident. what os you on?
22:50:46  <isaacs>first one to lgtm wins 7 internets: https://github.com/joyent/node/pull/5521
22:55:37  <tjfontaine>I'm mostly happy about all the deprecation
22:55:46  <trevnorris>isaacs: in process. mind glancing at 5520
22:57:48  <amartens>isaacs: love the idea, especially the comma-separated environment variable
22:59:20  <trevnorris>isaacs: lgtm. just had one nit
23:01:01  <tjfontaine>isaacs: also lgtm
23:03:50  <trevnorris>tjfontaine: any way tests could be run to find things like 5520?
23:04:42  <tjfontaine>I'm not sure what you mean
23:05:52  <tjfontaine>oh I'm on the wrong issue
23:05:53  <trevnorris>tjfontaine: there's a serious memory leak there that ended up crashing my machine. i'm trying to bisect it out right now. could, say, a pummel test be written to check mem usage and report if it's above X?
23:05:53  <tjfontaine>heh
23:06:08  <tjfontaine>I was on #5220 not #5520
23:06:11  <trevnorris>tjfontaine: problem is it's a race-y condition, so it depends on how quickly the server can process requests.
23:06:12  <trevnorris>oh, heh
23:06:37  <tjfontaine>trevnorris: there are some memory related tests in pummel and gc
23:06:43  <trevnorris>ok
23:06:58  <tjfontaine>one I know for sure is broken, but I haven't investigated yet
23:09:49  <piscisaureus_>hey isaacs
23:10:27  <piscisaureus_>when `npm ls` tells me
23:10:27  <piscisaureus_> npm ERR! invalid: debug@0.7.2 D:\supported-modules\node_modules\debug
23:10:27  <piscisaureus_> npm ERR! not ok code 0
23:10:27  <piscisaureus_>after it just (succesfully) installed that module?
23:10:27  <piscisaureus_>what does that mean?
23:16:17  * toothrchanged nick to toothrot
23:23:12  * timoxleyjoined
23:23:31  <amartens>trevnorris: regarding the maxresident, I'm on OS X here
23:24:01  <trevnorris>amartens: ok thanks.
23:24:20  <amartens>is there anything else I can do to get that displayed?
23:24:39  <amartens>or I could spin up an EC2 instance and test over there
23:25:05  <trevnorris>no worries.
23:25:58  * amartenskicks slightly non-useful OS
23:26:08  <trevnorris>well suck. 62a2142 can't run the test. and it was the last step in the bisect.
23:26:17  <trevnorris>amartens: might just need to pass an option.
23:26:48  <amartens>I'll check that in a few once I'm done this meeting
23:28:11  * paddybyersquit (Ping timeout: 260 seconds)
23:30:44  <amartens>ah, /usr/bin/time -lp on OS X
23:30:59  <amartens>trevnorris: 1110425600 maximum resident set size
23:31:09  <trevnorris>amartens: ok thanks.
23:31:25  <trevnorris>seems the BIO changes indutny introduced have the potential for memory leak.
23:31:33  <trevnorris>well, not memory leak exactly
23:31:50  <indutny>well...
23:31:57  <indutny>what are you doing?
23:32:06  <trevnorris>indutny: https://github.com/joyent/node/issues/5520
23:32:27  <trevnorris>indutny: sorry, didn't mean the BIO changes exactly.
23:32:36  <trevnorris>just something to do w/ crypto. =)
23:34:59  <trevnorris>indutny: just at a glance I'd think it has something to do with https://github.com/joyent/node/commit/62a2142#L0R283
23:35:13  <trevnorris>but have no real idea right now
23:35:14  <indutny>well, basically
23:35:18  <indutny>the ring isn't shrinking
23:35:39  <indutny>I think I could fix it
23:35:54  <indutny>like if it grows too much, it should shrink when its empty
23:36:28  <trevnorris>indutny: is that ring not tracked by v8?
23:36:35  <indutny>it isn't
23:36:56  <trevnorris>ah ok. that would explain why it didn't show up anywhere.
23:37:18  <trevnorris>now, have an idea why if the test is artificially slowed down it doesn't explode?
23:37:24  <trevnorris>seems to only happen under heavy load
23:38:20  <trevnorris>indutny: oh, so you're allocating a zero length Buffer then throwing memory onto it?
23:38:29  <indutny>no
23:38:33  <indutny>buffer has data embedded
23:38:41  <indutny>and that's not that buffer
23:38:49  <trevnorris>what's this line doing? Buffer* next = new Buffer();
23:38:50  <indutny>it has its own Buffer :)
23:40:15  <MI6>joyent/node: isaacs master * 896b2aa : util: Add debuglog, deprecate console lookalikes - http://git.io/SojPHA
23:40:45  <trevnorris>oh, wtf. somehow there's a bunch of "Buffer*" in node_crypto_bio.h, but that's never caused any problems in my Buffer branch.
23:40:59  <trevnorris>how is that possible? no build or tests have failed.
23:41:05  <trevnorris>(anyways, that's another issue)
23:42:15  <indutny>that's not that Buffers
23:42:20  <MI6>joyent/node: Timothy J Fontaine v0.10 * a846d93 : net: use timers._unrefActive for internal timeouts (+1 more commits) - http://git.io/OEOcFg
23:42:20  <trevnorris>...
23:42:27  <indutny>look at node_crypto_bio.h
23:42:37  <indutny>it has it's own Buffer class or struct
23:42:39  <indutny>whatever
23:42:54  <trevnorris>ah ok. so just a duplicate name. got it.
23:44:07  <trevnorris>indutny: well, seems that the objects are cleaned up eventually. how is the memory cleaned up?
23:47:39  <indutny>its cleaned up on destruction of ring :)
23:47:50  <indutny>so only after connection will be closed
23:47:59  <indutny>that's terrible thing, I know
23:48:08  <trevnorris>ok
23:48:29  * kevinswiberjoined
23:49:14  <trevnorris>oh, and the heavier the load the more simultaneous connections will be open.
23:49:33  <trevnorris>i mean. when I tried to run the test for 120 sec it filled my 8 GB of memory and swap
23:50:31  <MI6>nodejs-master: #227 FAILURE smartos-x64 (6/600) osx-x64 (3/600) linux-x64 (3/600) linux-ia32 (4/600) smartos-ia32 (3/600) http://jenkins.nodejs.org/job/nodejs-master/227/
23:50:58  <tjfontaine>stderr: fatal: Unable to look up github.com (port 9418) (nodename nor servname provided, or not known)
23:51:30  <tjfontaine>thank god we use a Dvcs just like a Cvcs
23:52:06  <trevnorris>lol