00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:00:45  * icarotquit (Remote host closed the connection)
00:01:13  * icarotjoined
00:05:07  * loladiroquit (Ping timeout: 246 seconds)
00:06:30  <isaacs>yeah
00:19:43  <isaacs>tjfontaine: oh!! no, this is a compeletely different bug.
00:19:46  <isaacs>tjfontaine: haha
00:19:57  <isaacs>tjfontaine: so, i found a case where the tls stream will hang
00:20:47  <tjfontaine>oh?
00:21:10  <tjfontaine>I have a test case that writes a ton and sends us into recursion death :)
00:21:24  <tjfontaine>but it's mostly just badly behaving code
00:24:24  <isaacs>tjfontaine: https://gist.github.com/isaacs/5854932
00:24:30  <isaacs>tjfontaine: if you remove the two read(0) calls, it hangs
00:24:38  <isaacs>tjfontaine: in v0.10, this fails worse.
00:24:48  <isaacs>but it does fail with my GH-5712 fixes
00:24:52  <isaacs>just not as big a diff
00:25:17  <isaacs>tjfontaine: which leads me to believe that the CryptoStream "socket" is emitting 'drain' early
00:25:35  <tjfontaine>ah
00:29:55  * qardquit (Quit: Leaving.)
00:31:29  <isaacs>ok, found the hangup
00:31:41  <isaacs>but still not getting the right number of bytes at the end.
00:31:52  <isaacs>so 'finish' is emitting before it's actually finished and consumed
00:31:56  * isaacscan't wiat for TLSSocket
00:32:02  <tjfontaine>ya
00:47:02  <tjfontaine>netstat -an | grep 443 | grep TIME_WAIT | wc -l
00:47:03  <tjfontaine>327
00:47:17  <tjfontaine>well how about that
00:48:02  <tjfontaine>that's pretty nasty
00:49:10  <tjfontaine>no wonder mark didn't like the default agent, isaacs is better-agent a module now?
00:50:40  <isaacs>tjfontaine: no, it's still just a pull request
00:50:46  <tjfontaine>right
00:58:04  * qardjoined
00:58:10  <isaacs>indutny: ping
00:58:16  <isaacs>indutny: found a weird tls bug
00:58:26  <tjfontaine>#2 :)
00:58:42  <tjfontaine>anyway, I'm going to commute home now, I'll bbs
00:59:13  <isaacs>ahhhh!!!!
00:59:18  <isaacs>ok... so:
00:59:20  <isaacs>tjfontaine: go home
00:59:21  <tjfontaine>ok
00:59:24  <tjfontaine>heh
00:59:25  <isaacs>tjfontaine: this'll be waiting when you get there
00:59:26  <isaacs>hurry!
00:59:28  <tjfontaine>k
00:59:39  * isaacswaiting 5 minutes so that tjfontaine doesn't get tempted and miss the boat ;)
01:01:11  <isaacs>so, IF we do this thing, where http.OutgoingMessage only emits 'finish' when the *socket* has accepted and successfully written the last bit of data, then that means that it'll still always fail for TLS. because, the "socket" from the OutgoingMessage's pov is actually a CryptoStream. That CryptoStream will emit 'finish' when its partner (the pair.encrypted side) has accepted the data. However, that *isn't* the time when the actual TCP socket has actually a
01:08:36  <TooTallNate>tjfontaine: boat?
01:08:55  <isaacs>TooTallNate: tjfontaine lives on the Island of Alamed
01:08:56  <isaacs>a
01:09:03  <isaacs>TooTallNate: it's only accessible by boat.
01:09:08  <isaacs>or bus, or you can walk there.
01:09:17  <isaacs>or drive
01:09:19  <TooTallNate>lol
01:09:21  <isaacs>it's pretty accessible, actually
01:09:33  <TooTallNate>so TJ takes the ferry?
01:10:10  <isaacs>yeah
01:10:32  <isaacs>so, i think for this 'finish isn't finished' issue, we relaly can't do it properly until 0.12.
01:10:39  <isaacs>we can make it less bad, but it'll still be bad for TLS
01:14:42  * amartens1quit (Quit: Leaving.)
01:20:18  * mikealquit (Quit: Leaving.)
01:20:29  <isaacs>unless, we have the TLS outgoing message wait for its cryptostream AND its cryptostream's socket to both drain
01:23:25  * icarotquit (Ping timeout: 268 seconds)
01:24:43  * st_lukequit (Remote host closed the connection)
01:28:35  * abraxasjoined
01:38:29  * icarotjoined
01:48:11  * dapquit (Quit: Leaving.)
02:03:19  * amartensjoined
02:05:37  <tjfontaine>isaacs: right, that all makes sense
02:06:20  <tjfontaine>TooTallNate: ya, the ferry is very relaxing, albeit slightly inconvenient at times :)
02:06:36  <TooTallNate>tjfontaine: i take the ferry too!
02:06:44  <TooTallNate>tjfontaine: to Larkspur though
02:06:48  <TooTallNate>not alameda :D
02:06:51  <tjfontaine>hehe
02:08:48  <tjfontaine>isaacs: I suppose it's not possible in 0.10 to dip down into the crypto side and see the pending outbound bio bytes
02:09:32  * TooTallNatequit (Quit: Computer has gone to sleep.)
02:27:45  * timoxleyquit (Quit: Computer has gone to sleep.)
02:28:44  * timoxleyjoined
02:31:02  * indexzerojoined
02:33:23  * brsonquit (Ping timeout: 255 seconds)
02:39:11  * dapjoined
02:50:22  * icarotquit (Ping timeout: 268 seconds)
03:10:13  * c4miloquit (Remote host closed the connection)
03:23:44  * dscape__quit (Ping timeout: 260 seconds)
03:32:10  * icarotjoined
03:32:25  <isaacs>tjfontaine: it's pretty gnarly
03:32:33  <tjfontaine>I figured
03:32:47  <isaacs>tjfontaine: the best you can get is whether the cleartext side has flushed it out to the encrypted side
03:33:04  <isaacs>tjfontaine: but then you don't know if the *socket* has flushed it, once the encrypted side reads it and writes to the tcp socket
03:33:39  <tjfontaine>ah, but you could find out if the encrypted bio has flushed to the socket
03:41:09  * dapquit (Quit: Leaving.)
03:42:20  * indexzeroquit (Quit: indexzero)
03:46:07  * TooTallNatejoined
03:57:45  <isaacs>right, but even then, , you'd have to track whether the last write returned true, or else 'drain' won't ever happen
03:57:49  <isaacs>tjfontaine: ^
03:58:02  <tjfontaine>right
04:04:45  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
04:05:02  <isaacs>it ends up being pretty complicated.
04:10:01  <tjfontaine>who needs tls
04:10:05  <tjfontaine>such a waste
04:10:24  <tjfontaine>if you're using it you must be trying to hide something
04:12:18  <isaacs>tjfontaine: exactly
04:12:26  <isaacs>just turning CPUs over more.
04:12:31  <isaacs>it's not environmental or green
04:13:08  <tjfontaine>WHY DO YOU HATE MOTHER EARTH BITCOINERS?!
04:13:09  <LOUDBOT>I'M GONNA WRITE A SIMCOPTER CLONE WITH PERL, XS, AND OGRE
04:13:15  <tjfontaine>heh
04:20:31  * TooTallNatejoined
04:21:35  * brsonjoined
04:25:50  * defunctzombie_zzchanged nick to defunctzombie
04:26:54  * brsonquit (Ping timeout: 246 seconds)
04:51:22  * inolenquit (Read error: No route to host)
05:06:53  * defunctzombiechanged nick to defunctzombie_zz
05:08:32  * defunctzombie_zzchanged nick to defunctzombie
05:09:45  * defunctzombiechanged nick to defunctzombie_zz
05:15:33  * Chip_Zeroquit (Ping timeout: 245 seconds)
05:22:54  * dscape__joined
05:26:43  * indexzerojoined
05:57:15  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
06:07:24  * qardquit (Quit: Leaving.)
06:12:15  * mikolalysenkoquit (Ping timeout: 246 seconds)
06:12:29  * mikolalysenkojoined
06:19:37  * mikealjoined
06:19:40  * mikealquit (Client Quit)
06:25:29  * mikealjoined
06:30:02  * `3E|Zzchanged nick to `3rdEden
06:45:49  * dominictarrjoined
06:49:13  * dominictarrquit (Client Quit)
06:51:26  * dominictarrjoined
06:52:13  * indexzeroquit (Quit: indexzero)
07:03:56  * csaohjoined
07:12:05  * csaohquit (Quit: csaoh)
07:16:20  * mikolalysenkoquit (Ping timeout: 246 seconds)
07:16:28  * Chip_Zerojoined
07:17:48  * timoxleyquit (Quit: Computer has gone to sleep.)
07:18:47  * amartensquit (Quit: Leaving.)
07:25:17  * timoxleyjoined
07:30:52  * csaohjoined
07:31:17  * dominictarrquit (Quit: dominictarr)
07:49:10  * amartensjoined
07:55:58  * amartensquit (Ping timeout: 256 seconds)
08:18:53  * st_lukejoined
08:23:16  * mikolalysenkojoined
08:27:59  * mikolalysenkoquit (Ping timeout: 255 seconds)
08:33:44  * st_lukequit (Remote host closed the connection)
08:34:03  * st_lukejoined
08:53:44  * amartensjoined
08:58:43  * amartensquit (Ping timeout: 264 seconds)
08:59:22  * st_lukequit (Remote host closed the connection)
09:00:20  * AvianFluquit (Remote host closed the connection)
09:04:10  * st_lukejoined
09:18:41  * timoxleyquit (Quit: Textual IRC Client: www.textualapp.com)
09:30:24  * icarotquit (Ping timeout: 240 seconds)
09:32:13  * timoxleyjoined
09:39:54  * bajtosjoined
09:54:02  * amartensjoined
09:56:24  * icarotjoined
09:57:10  * st_lukequit (Remote host closed the connection)
09:59:01  * amartensquit (Ping timeout: 276 seconds)
10:01:12  * bnoordhuisjoined
10:03:28  * icarotquit (Ping timeout: 256 seconds)
10:10:06  * abraxasquit (Remote host closed the connection)
10:12:33  * bajtosquit (Quit: bajtos)
10:14:42  <indutny>isaacs: hi
10:14:45  <indutny>what's up with tls?
10:23:15  <bnoordhuis>indutny: morning fedor. what are you working on
10:23:22  <indutny>morning ben
10:23:25  <indutny>I'm working on asm.rs
10:23:29  <indutny>assembler written in rust
10:23:31  <bnoordhuis>hah
10:23:33  <indutny>any other suggestions
10:23:34  <indutny>:)
10:23:41  <bnoordhuis>what machine language?
10:23:44  <indutny>I'm almost open to them
10:23:53  <indutny>bnoordhuis: right now ia64
10:23:58  <indutny>will do x86 later
10:24:00  <bnoordhuis>ah okay
10:24:01  <indutny>and arm
10:24:10  <indutny>then I'll be able to write ultimate JIT engine in rust
10:24:10  <bnoordhuis>i wrote a 6502 assembler in js this week :)
10:24:13  <indutny>oh
10:24:18  <indutny>nice
10:24:22  <bnoordhuis>apropos nothing
10:24:33  <bnoordhuis>what kind of jit are you going to write?
10:24:39  <indutny>erlang
10:24:46  <indutny>but that's just to test my engine
10:24:54  <indutny>it should be applicable to anything
10:24:56  <indutny>more or less
10:25:06  <bnoordhuis>you mean beam?
10:25:16  <indutny>yep
10:25:22  <bnoordhuis>nice
10:25:24  <indutny>btw, you should consider using rust for such stuff too: https://github.com/indutny/asm.rs/blob/master/src/x64/basic.rs
10:25:27  <indutny>matching looks really nice
10:25:41  <bnoordhuis>yeah, but they're still changing the language every now and again
10:25:49  <indutny>well, not that much
10:25:54  <indutny>and its simple to keep track of things
10:26:03  <indutny>just ask a couple of persons
10:26:32  <indutny>so… do you have some ideas?
10:26:33  <indutny>about node.js
10:27:02  <bnoordhuis>well, my grand scheme for storing pointers in nan-tagged doubles got foiled
10:27:17  <bnoordhuis>(that wasn't really my grand scheme)
10:27:31  <bnoordhuis>do you know why v8 doesn't store doubles unboxed on x64?
10:28:13  <bnoordhuis>as to node.js work, i guess there's a couple of outstanding bug reports and pull requests
10:28:21  <indutny>haha
10:28:32  <indutny>bnoordhuis: well, that's pretty much expected
10:28:39  <bnoordhuis>yeah :)
10:28:49  <bnoordhuis>btw, what's your opinion on adding a printf module to core?
10:28:57  <indutny>I'm ok with it
10:29:01  <bnoordhuis>okay, nice
10:29:09  <indutny>do you want to support "%.*s" ?
10:29:23  <indutny>and all the weirdness
10:29:24  <bnoordhuis>yeah, i suppose so. %*.*d as well
10:29:30  <bnoordhuis>i actually have that last one working
10:29:39  <indutny>nice
10:30:20  <indutny>so what do you need 6505 assembly for? :)
10:30:23  <indutny>err
10:30:25  <indutny>6502*
10:30:27  <bnoordhuis>ah
10:30:36  <bnoordhuis>i'm doing a talk later this week
10:30:46  <bnoordhuis>i wanted to demo a node-powered 6502 emulator
10:30:55  <bnoordhuis>but i don't think i have enough time to finish it up
10:31:06  <bnoordhuis>oh, when i say '6502 emulator' i mean 'nes emulator'
10:31:16  <bnoordhuis>the 6502 cpu emulator itself works
10:31:19  <indutny>transcompiling? :)
10:31:27  <indutny>its quite hard, actually
10:31:32  <indutny>I remember reading a couple of articles about it
10:31:37  <bnoordhuis>to do faithfully? yes, indeed
10:31:59  <bnoordhuis>my original plan was to emulate the physical hardware
10:32:09  <indutny>hehe
10:32:11  <indutny>so
10:32:21  <indutny>you need both assembly and disassembly for that
10:32:31  <bnoordhuis>yeah. i have both
10:32:37  <indutny>nice!
10:32:54  <indutny>so… it shouldn't be hard to do software-emulator
10:32:54  <bnoordhuis>re: emulating physical hardware, it's actually quite straining
10:33:13  <bnoordhuis>my core i7 can only do it at a slowdown of 250x
10:33:14  <indutny>why?
10:33:27  <indutny>:)
10:33:31  <bnoordhuis>that's 1/250th of ~1.33 mhz :)
10:33:38  <indutny>just schedule ticks :P
10:34:13  <bnoordhuis>well, i was thinking you can probably do it with a fpga
10:34:19  <bnoordhuis>but i digress
10:34:24  <bnoordhuis>okay, back to node
10:34:30  * csaohquit (Quit: csaoh)
10:34:32  <indutny>haha
10:34:43  <bnoordhuis>if you want to do something interesting
10:34:45  <indutny>running 6502 on fpga should be really fun
10:34:56  <bnoordhuis>there's still the issue of libuv's thread pool implementation
10:35:03  <bnoordhuis>and how it kind of sucks
10:35:30  <bnoordhuis>i did some work on an auto-scaling thread pool a few months ago
10:35:41  <bnoordhuis>but it's actually pretty hard to get good performance out of it
10:36:14  <bnoordhuis>i mean, the kind of performance that's okay overall, not just great on some benchmarks and completely tanking on others
10:36:23  <bnoordhuis>so, if you're interested, the job is yours
10:40:19  <indutny>hm...
10:40:22  <indutny>sounds pretty generic
10:40:26  <indutny>but I'll try to look into it :)
10:40:54  * csaohjoined
10:41:17  <bnoordhuis>indutny: generic?
10:41:24  <indutny>vague
10:41:47  <bnoordhuis>ah
10:41:52  <bnoordhuis>wait, let me link you to the code
10:42:56  <bnoordhuis>indutny: https://github.com/bnoordhuis/libuv/commits/threadpool-rework
10:43:01  * bajtosjoined
10:43:06  <bnoordhuis>man, 5 months ago already? time flies
10:44:06  <indutny>aaah
10:44:10  <indutny>you haven't finished it
10:44:16  <bnoordhuis>there's a bug in https://github.com/bnoordhuis/libuv/commit/9703ab44 btw, it always grows the threadpool (up to the limit) rather than only when there are no free workers
10:48:37  * bajtosquit (Quit: bajtos)
10:51:11  <MI6>joyent/node: Ben Noordhuis v0.10 * a0b6df0 : Revert "configure: fix cross-compilation host_arch_cc()" - http://git.io/KNG10Q
10:54:22  * amartensjoined
10:57:43  * timoxleyquit (Quit: Computer has gone to sleep.)
10:59:19  * amartensquit (Ping timeout: 264 seconds)
10:59:19  <indutny>hm...
11:01:59  <MI6>nodejs-v0.10: #258 UNSTABLE smartos-ia32 (1/592) linux-x64 (2/592) smartos-x64 (2/592) http://jenkins.nodejs.org/job/nodejs-v0.10/258/
11:06:18  * timoxleyjoined
11:10:52  <MI6>nodejs-v0.10-windows: #87 UNSTABLE windows-x64 (8/592) windows-ia32 (7/592) http://jenkins.nodejs.org/job/nodejs-v0.10-windows/87/
11:12:43  * rendarjoined
11:20:38  <indutny>bnoordhuis: hah http://en.wikipedia.org/wiki/Very_long_instruction_word
11:20:39  <indutny>that's nice
11:25:21  <bnoordhuis>indutny: vliw? yeah, well... look at (t)itanium
11:25:47  <bnoordhuis>you need very smart compilers to get the most mileage out of it
11:28:40  <indutny>yeah
11:31:25  <bnoordhuis>off to do the groceries, biab
11:35:41  * bnoordhuisquit (Ping timeout: 246 seconds)
11:43:15  * timoxleyquit (Quit: Computer has gone to sleep.)
11:45:18  * hzjoined
11:54:42  * amartensjoined
11:59:06  * amartensquit (Ping timeout: 246 seconds)
12:02:49  * timoxleyjoined
12:12:45  * timoxleyquit (Quit: Textual IRC Client: www.textualapp.com)
12:31:32  * bajtosjoined
12:46:33  <bajtos>ircretary: tell bnoordhuis pr#5713 is ready for your review. https://github.com/joyent/node/pull/5713
12:46:33  <ircretary>bajtos: I'll be sure to tell bnoordhuis
12:47:51  * dapjoined
12:55:02  * amartensjoined
12:59:13  * amartensquit (Ping timeout: 248 seconds)
13:11:13  * leonvvjoined
13:23:55  * c4milojoined
13:26:22  * defunctzombie_zzchanged nick to defunctzombie
13:39:00  * c4miloquit (Remote host closed the connection)
13:43:21  * mikolalysenkojoined
13:44:15  * piscisaureus_joined
13:48:06  * dominictarrjoined
13:55:24  * amartensjoined
13:59:54  * amartensquit (Ping timeout: 240 seconds)
14:05:06  * dominictarrquit (Quit: dominictarr)
14:11:47  * mmaleckiquit (*.net *.split)
14:11:47  * rvaggquit (*.net *.split)
14:13:21  * piscisaureus_quit (Ping timeout: 248 seconds)
14:13:23  * leonvvquit (Remote host closed the connection)
14:13:32  * rendar_joined
14:13:39  * papertigersjoined
14:14:08  * rendarquit (Read error: No buffer space available)
14:16:37  * mmaleckijoined
14:16:37  * rvaggjoined
14:27:50  * c4milojoined
14:30:26  * bajtosquit (Quit: bajtos)
14:30:58  * mmaleckiquit (*.net *.split)
14:30:58  * rvaggquit (*.net *.split)
14:34:48  <indutny>ircretary: tell bnoordhuis https://raw.github.com/6502/js6502/master/6502.js
14:34:48  <ircretary>indutny: I'll be sure to tell bnoordhuis
14:35:37  * mmaleckijoined
14:35:37  * rvaggjoined
14:40:25  * rendarjoined
14:43:24  * kellabyte2joined
14:47:19  * bnoordhuisjoined
14:49:18  * rendar_quit (*.net *.split)
14:49:19  * kellabytequit (*.net *.split)
14:50:21  * saghulquit (Ping timeout: 264 seconds)
14:50:54  * kenperkinsquit (Read error: Connection reset by peer)
14:50:55  * saghuljoined
14:51:38  * kenperkinsjoined
14:52:51  * othiym23quit (Ping timeout: 244 seconds)
14:53:19  * mikealquit (Ping timeout: 264 seconds)
14:54:10  * MI6quit (Ping timeout: 273 seconds)
14:54:39  * othiym23joined
14:54:57  * kellabyte2changed nick to kellabyte
14:55:06  * kellabytequit (Changing host)
14:55:06  * kellabytejoined
14:55:39  * amartensjoined
14:57:47  * AvianFlujoined
15:00:30  * amartensquit (Ping timeout: 268 seconds)
15:13:21  <bnoordhuis>no call today?
15:13:30  <bnoordhuis>or is it in 45 minutes?
15:14:58  * groundwaterjoined
15:16:15  * c4miloquit (Remote host closed the connection)
15:17:13  <tjfontaine>unclear
15:19:05  <bnoordhuis>okay. i'll assume the latter for now
15:19:24  * mikealjoined
15:22:02  <indutny>unclear
15:24:22  <bnoordhuis>indutny: re: that 6502 jit, how does it account for self-modifying code or instruction timing?
15:24:58  <indutny>that's a good question indeed
15:25:10  <indutny>I was thinking about interesting trick, btw
15:25:18  <indutny>custom fs
15:25:23  <indutny>and mmap'ed file from it
15:25:32  <indutny>and running executable from it
15:25:43  <indutny>but it might be approached in other way too
15:25:48  <indutny>probably much faster
15:25:50  <bnoordhuis>custom fs?
15:26:05  <indutny>kernel module
15:26:19  <indutny>but seems to be too much for 6502 emulation
15:28:28  <indutny>yeah
15:28:39  <indutny>it should be simplier to just translate memory stores into runtime calls
15:28:44  <indutny>or
15:28:47  <indutny>into write-barriers
15:29:18  <bnoordhuis>yeah. checked stores
15:29:23  <bnoordhuis>it doesn't seem to do that though
15:30:54  <bnoordhuis>still, looks pretty neat
15:33:11  * dapquit (Quit: Leaving.)
15:41:02  * toothrjoined
15:41:47  * mikealquit (Quit: Leaving.)
15:41:58  * roxlu_joined
15:43:07  <isaacs>call in 15 minutes
15:44:31  * niska`joined
15:46:01  <tjfontaine>ok
15:46:26  * Denmadquit (*.net *.split)
15:46:26  * toothrotquit (*.net *.split)
15:46:27  * rjequit (*.net *.split)
15:46:27  * roxluquit (*.net *.split)
15:46:27  * niskaquit (*.net *.split)
15:50:54  <indutny>call
15:50:57  <indutny>in 10 minutes
15:52:32  <isaacs>yes
15:53:08  * piscisaureus_joined
15:53:28  <isaacs>indutny:
15:53:28  <isaacs>=== release test-tls-client-destroy-soon ===
15:53:28  <isaacs>Path: simple/test-tls-client-destroy-soon
15:53:29  <isaacs>Command: out/Release/node /Users/isaacs/dev/js/node-master/test/simple/test-tls-client-destroy-soon.js
15:53:31  <isaacs>--- TIMEOUT ---
15:53:32  <LOUDBOT>WE SHOULD DO CHARADES OVER IRC
15:53:42  <indutny>platform?
15:53:46  <isaacs>indutny: os x
15:53:48  <isaacs>indutny: master
15:54:05  <indutny>nice!
15:54:05  <indutny>:)
15:54:31  <indutny>how you did it? :)
15:54:35  <isaacs>indutny: When i was investigating the 5712 thing, I found a case where the TLS crypto streams will hang
15:54:40  <isaacs>indutny: running `make test` :)
15:54:51  <indutny>its related to finish event?
15:54:57  <isaacs>indutny: yes, i think so
15:55:08  <isaacs>indutny: but, what's happening is that it writes a lot of data, but gets stuck
15:55:15  <isaacs>i have a small fix, one se
15:55:15  <isaacs>c
15:55:22  <indutny>one sec
15:55:25  <indutny>let me build it
15:56:04  * amartensjoined
15:56:17  <indutny>odd
15:56:21  <indutny>seems to be working for me
15:58:28  * Domenic_quit (Ping timeout: 268 seconds)
15:58:51  <trevnorris>the time change is throwing me off, did I miss today's call?
15:58:54  * c4milojoined
15:59:20  * Raynosquit (Ping timeout: 256 seconds)
15:59:50  <tjfontaine>not yet
16:00:15  * amartensquit (Ping timeout: 246 seconds)
16:00:43  <isaacs>sorry, skype is complaining
16:01:16  * bajtosjoined
16:02:09  * Raynosjoined
16:08:43  * brsonjoined
16:10:05  * mikealjoined
16:12:51  * Denmadjoined
16:12:51  * rjejoined
16:13:58  * dapjoined
16:14:27  * bajtosquit (Quit: bajtos)
16:14:40  * creationix_joined
16:15:11  * bajtosjoined
16:19:53  * creationixquit (Ping timeout: 264 seconds)
16:19:56  * creationix_changed nick to creationix
16:25:17  <isaacs>tjfontaine, indutny, trevnorris, bnoordhuis, piscisaureus_: ping
16:25:20  <tjfontaine>pogn
16:25:31  <indutny>pong
16:25:33  <isaacs>piscisaureus_: so, the tls_wrap stuff DID already land.
16:25:34  <piscisaureus_>isaacs: if you send an invite with google you can attach a hangout link
16:25:44  <piscisaureus_>isaacs: yes. So let's revert it?
16:26:18  <isaacs>piscisaureus_: and it's a bit of a disaster in windows
16:26:47  <piscisaureus_>isaacs: I'm not advocating for windows really but we can't really always turn windows into a half-broken state
16:26:49  <isaacs>tjfontaine: how much of a "disaster" is a "diaster?
16:26:54  <piscisaureus_>only to discover that just before the 0.12 release
16:27:00  <isaacs>i agree.
16:27:10  <tjfontaine>well, we're around 20 test failures, which is double than 0.10 :)
16:27:17  <isaacs>otoh, if it's reasonable to fix, then we should fix it
16:27:19  <piscisaureus_>isaacs: or we would cease to officially drop windows support
16:27:20  <isaacs>rather than revert it
16:27:33  <piscisaureus_>ER, or we would cease to officially support windows
16:27:50  <isaacs>that would be a bigger disaster, for sure?
16:28:11  <piscisaureus_>well I don't know. It's just about how serious are we :)
16:28:28  <piscisaureus_>I would be not for it I guess
16:28:53  <isaacs>i am not for dropping windows support
16:29:02  <isaacs>and, in fact, 10 failing tests in 0.10 is way too many
16:29:10  <isaacs>we have a zero-tolerance policy for test failures on unix
16:29:15  <isaacs>that's what "support" means
16:29:16  <bnoordhuis>fwiw, i wouldn't want to drop windows support either
16:29:24  <bnoordhuis>though i also don't really want to maintain it :)
16:29:44  <isaacs>i think if we all spent time working on it, we could get it much better.
16:29:57  <isaacs>that's what we did back in 0.6 -> 0.8
16:30:18  <isaacs>of course, we also had igor and bert was full time on it
16:30:31  <isaacs>adn now we have 20% of scott, and 20% of bert.
16:30:37  <tjfontaine>ya
16:30:56  <isaacs>maybe we need some bountysource action
16:32:47  <tjfontaine>I mean, some platform installer for microsoft is still installing 0.6.20 :/
16:32:58  <isaacs>indutny: i know it's not fun. would you be willing to take a look at the new tls failures on windows?
16:33:09  <indutny>sure
16:33:18  <isaacs>i'll email claudio and glenn and scott, see if we can make it more apparent that things are on fire.
16:33:41  <isaacs>reverting tls_wrap feels like failure. that would suck.
16:33:47  <isaacs>if it can't be fixed, i mean, we'll do it.
16:34:02  <indutny>surely yes
16:34:05  <indutny>what's 0.12 estimate?
16:34:29  <tjfontaine>it's just a stream stall, basically at some point we're not either drain'ing or reading anymore
16:34:57  <isaacs>tjfontaine: the test failures on master/winodws?
16:35:14  <isaacs>indutny: let's get windows stable with tlswrap, land trevors' slab allocator stuff, and the http keepalive stuff, and call it 0.12
16:35:18  <tjfontaine>ya, I had spent a little time on it, basically we stop reading at some point
16:35:24  <isaacs>indutny: or at least, call it feature complete, and then work on stability
16:35:25  <indutny>ok
16:35:30  * qardjoined
16:36:01  <isaacs>bnoordhuis, piscisaureus_, trevnorris, tjfontaine, indutny: How dows that plan sound to you? (genuinely asking, if you think this is a bad strategy, please say so)
16:36:08  <indutny>good
16:36:49  <tjfontaine>I am ok with it
16:37:38  <trevnorris>isaacs: does the "stuff" include the fix in http perf regression?
16:37:44  <isaacs>trevnorris: yes
16:37:51  <trevnorris>ok. them i'm good w/ it
16:38:19  * defunctzombiechanged nick to defunctzombie_zz
16:38:28  <piscisaureus_>isaacs: sounds good to me.
16:38:45  <piscisaureus_>isaacs: I guess it'll take some time to fix things?
16:38:55  <isaacs>piscisaureus_: yeah
16:39:02  <isaacs>i have no idea what the timeline looks like, really
16:39:19  <isaacs>piscisaureus_: windows is important in the enterprise, right? what's the likelihood of SL hiring a windows node dev?
16:39:32  * mikealquit (Quit: Leaving.)
16:39:33  <isaacs>piscisaureus_: or leaning on contacts at microsoft to get them to give us more resources?
16:39:51  <isaacs>piscisaureus_: or is that a crazy line of thinking, and actually not something SL cares about?
16:40:24  <bnoordhuis>i think we'd be open to it but it's not that easy to find competent windows system programmers
16:40:37  <piscisaureus_>isaacs: SL cares about it, but there's more to it
16:40:56  <isaacs>of course
16:42:20  <bnoordhuis>isaacs: btw, re v0.12, i kind of promised github to land their multi-context stuff in 0.12
16:42:39  <isaacs>bnoordhuis: oh, ok
16:42:44  <isaacs>bnoordhuis: we can get that in, too.
16:42:54  <bnoordhuis>okay, nice
16:43:08  <isaacs>bnoordhuis: but if everything else is done, and that's the only hold up, it's going to suck to delay for that.
16:43:15  * amartensjoined
16:43:18  <isaacs>bnoordhuis: i think you've probably got plenty of time, though. this other stuff will take a while.
16:43:40  * defunctzombie_zzchanged nick to defunctzombie
16:44:11  <bnoordhuis>fwiw, i'm downloading windows 7 as we speak :)
16:44:27  <trevnorris>heh
16:44:39  <bnoordhuis>btw, do we still support xp?
16:44:44  <tjfontaine>I don't.
16:44:45  <tjfontaine>:P
16:44:56  <bnoordhuis>i noticed this week that there's a ton of libuv failures on xp
16:45:02  <bnoordhuis>i didn't have the heart to run node's
16:45:54  <piscisaureus_>I'm fine with dropping XP but it makes no sense unless we also drop server 2k3
16:46:19  <piscisaureus_>in fact it doesn't help, running on xp is most likely just exposing some bugs that are there anyway
16:46:30  <bnoordhuis>right
16:46:34  <piscisaureus_>the code is only minimally different
16:47:52  <bnoordhuis>i won't make jokes about how MS slapped a new skin on it and called it vista
16:48:02  <bnoordhuis>because i know that's not what you meant :)
16:50:18  <indutny>xp
16:50:26  <indutny>lets drop XP
16:50:35  <indutny>there so many stuff that has been created since it
16:54:32  * EhevuTovjoined
16:55:12  <indutny>rebooting to windows is a good reason to download skyrim
16:57:11  * csaohquit (Quit: csaoh)
17:00:04  * TooTallNatejoined
17:03:59  * mikolalysenkoquit (Ping timeout: 246 seconds)
17:06:00  * brsonquit (Ping timeout: 260 seconds)
17:06:23  * trevnorris&
17:06:23  <LOUDBOT>YOU SHOULD BE USING CHROME YOU SILLY ASS
17:07:11  * c4miloquit (Remote host closed the connection)
17:09:55  * mikolalysenkojoined
17:10:47  * brsonjoined
17:13:36  * Domenic_joined
17:15:29  * amartens1joined
17:16:20  * mikealjoined
17:16:25  * \toothrotjoined
17:18:34  * mikeal1joined
17:18:37  * mikealquit (Read error: Connection reset by peer)
17:20:28  * saghul_joined
17:21:48  * amartensquit (*.net *.split)
17:21:48  * toothrquit (*.net *.split)
17:21:48  * saghulquit (*.net *.split)
17:21:49  * dlmanningquit (*.net *.split)
17:21:53  * saghul_changed nick to saghul
17:22:17  * Benvie_quit (Ping timeout: 252 seconds)
17:22:54  * Benvie_joined
17:29:24  * dlmanningjoined
17:34:28  * brsonquit (Read error: Operation timed out)
17:34:41  * brsonjoined
17:35:52  * amartens1changed nick to amartens
17:41:42  * loladirojoined
17:42:21  * amartensquit (Ping timeout: 240 seconds)
17:44:01  * amartensjoined
17:54:57  * c4milojoined
17:56:29  * c4miloquit (Remote host closed the connection)
18:03:37  * groundwaterquit (Quit: groundwater)
18:05:44  * brsonquit (Ping timeout: 260 seconds)
18:05:44  * piscisaureus_quit (Ping timeout: 260 seconds)
18:10:44  * brsonjoined
18:19:46  * bajtosquit (Quit: bajtos)
18:27:04  * icarotjoined
18:36:21  <ryah>dropping windows support?
18:36:43  <ryah>i hope this is some sort of joke
18:38:02  * ryahquit (Quit: leaving)
18:38:11  <isaacs>ryah: it's not a joke, it's a statement of a problematic state we're in
18:38:16  <tjfontaine>he's already gone
18:38:21  <isaacs>ah, indeed.
18:47:00  * c4milojoined
18:47:02  <isaacs>merging in 0.10 to master
18:47:21  <isaacs>can't seem to get that TLS hang to happen again
18:47:22  * MI6joined
18:47:23  <isaacs>it's super weird
18:47:31  <isaacs>wb, MI6
18:48:57  <tjfontaine>it's a frustrating world
18:49:52  <isaacs>everythign is terrible
18:49:57  <isaacs>brb, iterm2 restarting
18:56:13  * bnoordhuisquit (Read error: Connection reset by peer)
18:57:02  * piscisaureus_joined
18:58:53  <MI6>nodejs-master: #272 UNSTABLE smartos-x64 (4/606) linux-x64 (2/606) smartos-ia32 (2/606) osx-x64 (1/606) http://jenkins.nodejs.org/job/nodejs-master/272/
19:02:42  * dominictarrjoined
19:03:59  * c4miloquit (Remote host closed the connection)
19:04:58  * mikolalysenkoquit (Ping timeout: 245 seconds)
19:07:51  * c4milojoined
19:10:31  * bnoordhuisjoined
19:14:37  <MI6>nodejs-master-windows: #81 UNSTABLE windows-x64 (18/606) windows-ia32 (20/606) http://jenkins.nodejs.org/job/nodejs-master-windows/81/
19:17:38  * mcavagejoined
19:21:15  * trapitojoined
19:23:38  * defunctzombiechanged nick to defunctzombie_zz
19:24:57  * dominictarrquit (Quit: dominictarr)
19:42:53  * defunctzombie_zzchanged nick to defunctzombie
19:43:14  * dominictarrjoined
19:45:23  * brsonquit (Ping timeout: 240 seconds)
19:52:59  * mikeal1quit (Quit: Leaving.)
20:06:09  <bnoordhuis>piscisaureus_: is fs_file_loop expected to fail on win7?
20:06:30  * HenryRjoined
20:08:04  * HenryRquit (Client Quit)
20:13:11  <kellabyte>going to be trying to get libuv up into 1Gbps tonight, wish me luck :P
20:14:08  * AvianFluquit (Remote host closed the connection)
20:16:47  <bnoordhuis>kellabyte: let me know how that works out :)
20:17:08  <bnoordhuis>i'd be interested in the perf data (perf the tool) if you can get it
20:18:45  * st_lukejoined
20:21:19  <kellabyte>bnoordhuis: actually just found out the I/O boundary is 800Mbps so I can't get to 1Gbps on this network, I was able to hit 714Mbps last night
20:21:46  <kellabyte>I can collect perf data if you'd like running that same test
20:21:58  <kellabyte>there's a bunch of my code on top though
20:23:12  <kellabyte>thats 714Mbps with http_parser blasting through 560,000 pipelined requests/second
20:23:24  * mikealjoined
20:25:03  <bnoordhuis>hah, nice
20:25:38  * mikealquit (Read error: Operation timed out)
20:26:11  <isaacs>kellabyte: nice :)
20:26:18  <tjfontaine>excellent
20:27:56  <bnoordhuis>piscisaureus_: i have 16-17 failing libuv tests on win7. is that expected?
20:28:14  <tjfontaine>what's different about win7 vs 2k8 that confuses me
20:28:48  <kellabyte>thanks for the wonderful projects, it's so fun to learn C with such great libraries :)
20:28:49  <bnoordhuis>you mean everything passes on 2k8?
20:29:03  <tjfontaine>I mean on v0.10 there's only the thread tests outstanding
20:29:19  <bnoordhuis>oh, this is libuv master. sorry, should've specified that
20:29:54  <tjfontaine>still rougly the same, sometimes hrtimer is flakey everywhere though, just threadpool
20:30:00  <tjfontaine>http://jenkins.nodejs.org/job/libuv-master/label=windows/lastCompletedBuild/tapTestReport/
20:30:30  <bnoordhuis>hm, weird
20:30:52  <bnoordhuis>i have a ton of failing uv_fs_event_* tests
20:31:11  <tjfontaine>are you running with elevated privileges? that may be a difference
20:31:19  <tjfontaine>or not as the case may be
20:31:32  <bnoordhuis>err, how do i find out?
20:32:05  <bnoordhuis>if it's anything like gsudo, i didn't enter any password :)
20:32:20  <tjfontaine>well the account type you made is "adminstrator" of sorts
20:32:32  <tjfontaine>you're not running as a power-user or regular-user
20:33:34  <bnoordhuis>right. it's an admin account
20:34:09  <bnoordhuis>those are still restricted by default, right?
20:34:35  <tjfontaine>somewhat, but shouldn't be in such a way where fs events fail, at least not in my experience
20:36:26  <piscisaureus_>bnoordhuis: fs_event is not supposed to fail on windows...
20:36:58  <piscisaureus_>bnoordhuis: or are you using shared folders or something?
20:38:59  * st_lukequit (Remote host closed the connection)
20:40:27  <bnoordhuis>piscisaureus_: yes
20:40:45  <bnoordhuis>i take it that doesn't work?
20:44:31  * txdv_joined
20:45:11  * txdvquit (Ping timeout: 246 seconds)
20:49:18  * indexzerojoined
20:50:46  <piscisaureus_>bnoordhuis: I don't know, it wouldn't surprise me
20:50:53  <piscisaureus_>bnoordhuis: to windows, this is a network drive
20:51:14  <piscisaureus_>bnoordhuis: and you can't do fs events on network shares except some exceptions
20:54:00  <bnoordhuis>piscisaureus_: right, on a regular drive i'm down to the usual suspects and fs_file_loop and ipc_send_recv_tcp
20:58:23  * hzquit
21:02:00  * brsonjoined
21:04:54  <piscisaureus_>bnoordhuis: ipc_send_recv_tcp might be genuine
21:05:10  <piscisaureus_>bnoordhuis: fs_file_loop dunno - waht windows is thath?
21:05:16  <piscisaureus_>what and that
21:05:21  <piscisaureus_>how hard can it be?
21:05:52  <tjfontaine>win7 is what he's on
21:07:55  * mikealjoined
21:12:13  * rendarquit
21:15:30  * EhevuTov_joined
21:17:53  * EhevuTovquit (Ping timeout: 245 seconds)
21:23:54  * icarotquit (Ping timeout: 240 seconds)
21:29:43  <indutny>tjfontaine: yikes!
21:29:45  <indutny>I've fixed it
21:29:51  <indutny>tadam
21:29:55  <indutny>tls_wrap on windows
21:30:07  <indutny>also, probably an osx issue that isaacs has reported
21:34:32  * st_lukejoined
21:35:14  <indutny>ircretary: tell isaacs to check out https://github.com/joyent/node/pull/5746 on osx
21:35:14  <ircretary>indutny: I'll be sure to tell isaacs
21:35:40  <tjfontaine>oh?
21:36:12  <tjfontaine>hehe you really want my name to have a u in it :)
21:39:56  * AvianFlujoined
21:40:11  * indexzeroquit (Quit: indexzero)
21:41:36  * papertigersquit (Quit: papertigers)
21:53:26  <bnoordhuis>piscisaureus_: what are uv_utf8_to_utf16 and uv_utf16_to_utf8 doing in uv-win.h?
21:54:07  <piscisaureus_>bnoordhuis: being tested iirc
21:54:25  <piscisaureus_>bnoordhuis: dunno. did I putblame?
21:54:34  <bnoordhuis>no
21:55:09  <bnoordhuis>ah, uv_utf8_to_utf16 is indeed being used in a test
21:55:19  <bnoordhuis>the other one isn't however
21:55:56  <piscisaureus_>bnoordhuis: btw - those functions are deprecated
21:56:10  <piscisaureus_>bnoordhuis: use WideCharToMultiByte and MultiByteToWideChar directly
21:56:38  <piscisaureus_>bnoordhuis: I think I removed at most places but not everywhere
21:56:45  <bnoordhuis>i'm not using them, just noticed they're declared in a public(ish) header
21:57:01  <bnoordhuis>okay, maybe i'll clean them up as an encore
21:57:21  <piscisaureus_>cool. you have my sword
21:59:12  <indutny>ah
21:59:18  <indutny>tjfontaine: sorry
21:59:37  <indutny>its just he fountain
22:00:06  <indutny>bnoordhuis: have you seen my PR?
22:00:12  <indutny>bnoordhuis: does it looks good to you https://github.com/joyent/node/pull/5746?
22:00:27  <tjfontaine>hehe, it may be that my ancestors were La Fountain, but we americanized it in the early 1900s
22:00:43  <indutny>haha
22:00:49  <indutny>probably
22:01:01  <tjfontaine>that's the oral history of my family anyway
22:01:14  * bradleymeckjoined
22:01:16  <indutny>so
22:01:20  <bnoordhuis>indutny: second please
22:01:20  <indutny>jenkins seems to be feeling bad
22:01:24  <indutny>1
22:01:26  <indutny>...
22:01:27  <indutny>0
22:01:27  * AvianFluquit (Read error: Connection reset by peer)
22:02:04  * AvianFlujoined
22:04:58  <indutny>ok
22:05:00  <indutny>time to sleep
22:05:01  <indutny>ttyl
22:05:39  <MI6>libuv-master-gyp: #63 UNSTABLE windows-x64 (3/190) windows-ia32 (3/190) smartos-ia32 (2/189) smartos-x64 (2/189) http://jenkins.nodejs.org/job/libuv-master-gyp/63/
22:07:09  <bnoordhuis>sleep tight fedor
22:09:39  * hzjoined
22:09:45  * hzquit (Remote host closed the connection)
22:18:15  * mcavagequit (Remote host closed the connection)
22:19:10  * mmaleckiquit (*.net *.split)
22:19:10  * rvaggquit (*.net *.split)
22:37:19  <bnoordhuis>piscisaureus_: https://github.com/joyent/libuv/blob/master/src/win/tcp.c#L53 <- is using errno there a bug? i thought you had to call WSAGetLastError()
22:50:44  <piscisaureus_>bnoordhuis: yes. bug
22:51:23  <piscisaureus_>bnoordhuis: also a couple of lines down
22:52:25  <MI6>libuv-master-gyp: #64 UNSTABLE windows-x64 (3/190) windows-ia32 (4/190) smartos-ia32 (2/189) smartos-x64 (2/189) http://jenkins.nodejs.org/job/libuv-master-gyp/64/
22:52:44  * rvaggjoined
22:55:06  * mcavagejoined
22:59:12  * c4miloquit (Remote host closed the connection)
23:02:33  * mmaleckijoined
23:04:26  <MI6>joyent/libuv: Ben Noordhuis master * 30a8b44 : windows: use WSAGetLastError(), not errno - http://git.io/FkpWwg
23:06:04  <tjfontaine>should that also be on the v0.10 branch?
23:06:19  <piscisaureus_>it goes all the way back to 0.6 :)
23:06:41  <bnoordhuis>i'll backport it to v0.8 and v0.10
23:06:49  <bnoordhuis>because i'm nice like that
23:07:22  <tjfontaine>such a good guy
23:08:18  <MI6>libuv-master: #126 UNSTABLE windows (5/190) smartos (2/189) http://jenkins.nodejs.org/job/libuv-master/126/
23:09:46  <MI6>joyent/libuv: Ben Noordhuis v0.10 * c8c775b : windows: use WSAGetLastError(), not errno - http://git.io/GJqZbg
23:09:48  <MI6>joyent/libuv: Ben Noordhuis v0.8 * 0865a6f : windows: use WSAGetLastError(), not errno - http://git.io/jKmFHA
23:12:33  <MI6>libuv-v0.8: #23 UNSTABLE smartos (5/158) osx (1/158) linux (3/158) http://jenkins.nodejs.org/job/libuv-v0.8/23/
23:12:55  * st_luke_joined
23:14:46  <isaacs>bnoordhuis: what, no 0.6 backport? ;P
23:14:52  <MI6>libuv-v0.10: #99 UNSTABLE windows (3/187) smartos (2/186) http://jenkins.nodejs.org/job/libuv-v0.10/99/
23:15:42  * st_lukequit (Ping timeout: 252 seconds)
23:16:51  <MI6>libuv-v0.10-gyp: #62 UNSTABLE smartos-x64 (2/186) smartos-ia32 (2/186) linux-ia32 (2/186) windows-x64 (4/187) osx-ia32 (2/187) linux-x64 (2/186) windows-ia32 (16/187) http://jenkins.nodejs.org/job/libuv-v0.10-gyp/62/
23:19:27  * mikolalysenkojoined
23:26:16  * stagasjoined
23:28:27  <MI6>libuv-node-integration: #119 UNSTABLE linux-x64 (1/606) smartos-x64 (7/606) smartos-ia32 (11/606) http://jenkins.nodejs.org/job/libuv-node-integration/119/
23:31:20  * AvianFluquit (Read error: Connection reset by peer)
23:31:29  <MI6>libuv-node-integration: #120 FAILURE http://jenkins.nodejs.org/job/libuv-node-integration/120/
23:31:58  * AvianFlujoined
23:32:20  * loladiroquit (Quit: loladiro)
23:45:53  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
23:46:27  <MI6>libuv-node-integration: #121 FAILURE linux-x64 (1/592) smartos-x64 (2/592) smartos-ia32 (1/592) linux-ia32 (1/592) http://jenkins.nodejs.org/job/libuv-node-integration/121/
23:55:16  * Benvie_quit
23:55:49  * st_luke_quit (Remote host closed the connection)
23:56:01  * st_lukejoined
23:59:51  * mcavagequit (Remote host closed the connection)