00:00:01  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:01:28  * dapquit (Quit: Leaving.)
00:01:58  * dapjoined
00:10:09  * indexzeroquit (Quit: indexzero)
00:21:06  * paddybyersquit (Ping timeout: 276 seconds)
00:22:26  * qmxchanged nick to qmx|away
00:25:12  * TooTallNatejoined
00:29:36  * indexzerojoined
00:30:58  * EhevuTovquit (Quit: This computer has gone to sleep)
00:32:11  * hzquit
00:32:39  <isaacs>sblom: max ogden says you used to work at a startup in portalnd and do cool stuff with puzzles and he knows you
00:32:46  <isaacs>sblom: you should probably lurk in #Node.js :)
00:38:34  * lohkeypart
00:38:51  * lohkey_joined
00:38:51  * lohkey_quit (Client Quit)
00:47:47  * dapquit (Quit: Leaving.)
01:12:02  * mjr_quit (Quit: mjr_)
01:15:50  <sblom>isaacs: Cool, and roger on the don't-push-unless... stuff. Force pushes _always_ make me feel bad. ;-p
01:21:51  * trevnorrisquit (Quit: Leaving)
01:24:34  <sblom>isaacs: Yeah, Max and I used to bump into each other. The Portland tech scene has/had all sorts of cool folks in it.
01:24:44  <isaacs>nice
01:24:47  <isaacs>small world :)
01:24:52  <sblom>yup
01:25:36  <isaacs>you're also friends with my coworker laurel reitmann, i'm told.
01:25:46  <isaacs>or foafs at least.
01:31:12  <sblom>Actual first-order friends with her. She was my wife's maid of honor. :)
01:38:00  * stagasquit (Ping timeout: 256 seconds)
01:52:47  * pooyaquit (Quit: pooya)
01:59:31  * pooyajoined
02:01:14  <isaacs>ah, right
02:01:22  <isaacs>sblom: yes, i remember it was something about a wedding
02:01:42  <isaacs>tracking down performance issues is such an addictive passtime
02:05:55  * TooTallNatequit (Quit: Computer has gone to sleep.)
02:08:17  * pooyaquit (Quit: pooya)
02:32:04  * bnoordhuisquit (Ping timeout: 246 seconds)
02:36:33  <isaacs>So..... I think I broke something. https://docs.google.com/spreadsheet/ccc?key=0AganzoeqkiHddE9uRlA5WS0xSF9NTGN1QS0zZ3FRSWc#gid=1
02:36:38  <isaacs>streams2 is faster now.
02:37:06  <isaacs>my streams1 branch has the same v8 and libuv and src/, just the pre-streams2 lib/*.js
02:37:23  <isaacs>and it got slower somehow, while streams2 has gotten faster.
02:37:33  <tjfontaine>which version of v8, the latest upgrade?
02:37:43  * pieternquit (Ping timeout: 244 seconds)
02:40:00  <isaacs>tjfontaine: yeah
02:40:04  <isaacs>i'm assuming that i broke something
02:40:25  <isaacs>trying it a different way: branching off of master, and checking out the lib/*.js from the pre-streams2 merge
02:40:35  <isaacs>there's no way that it would be 2% slower.
02:40:41  <isaacs>unless i'm a wizard. but i don't think i am.
02:41:57  <tjfontaine>it's not impossible for a v8 regression to affect streams1, while overall boosting streams2 performance, maybe you *are* a wizard
02:43:45  <isaacs>hm. yeah.
02:43:55  <isaacs>i think the v8 update broke streams1
02:43:59  <isaacs>i'm getting a lot of errors.
02:44:06  <isaacs>probably we changed some interface in src/* actually
02:45:48  <isaacs>ok, so, branching off of master, and then checking out the streams1 lib/, though it fails LOTS of tests, is a bit faster on http_simple it would seem
02:46:32  <isaacs>or not.. i don't know. too close to tell.
02:46:44  <isaacs>it's +- 1% at this point, kinda going back and forth.
02:46:51  <isaacs>this is *very* encouraging, though
02:47:07  <tjfontaine>quite
02:47:22  <isaacs>once my ephemeral ports get out of tcp_wait, i'll run it again on streams 2 and get a closer comparison
02:55:09  * jmar777joined
02:55:59  <isaacs>ok, definitely getting reproducible results here. streams2 is faster than the pre-streams2 js on the current src/ and deps/
02:56:02  <isaacs>https://docs.google.com/spreadsheet/ccc?key=0AganzoeqkiHddE9uRlA5WS0xSF9NTGN1QS0zZ3FRSWc#gid=1
02:56:41  <isaacs>that's with a very slight change to add this._connecting=false into the Socket constructor, and inlining the logic of the _dump() method.
02:57:07  <isaacs>ircretary: tell piscisaureus to look at https://docs.google.com/spreadsheet/ccc?key=0AganzoeqkiHddE9uRlA5WS0xSF9NTGN1QS0zZ3FRSWc#gid=1
02:57:07  <ircretary>isaacs: I'll be sure to tell piscisaureus
02:57:18  * TooTallNatejoined
02:59:01  <tjfontaine>isaacs: definitely good news then
03:04:16  * lohkeyjoined
03:07:47  * TooTallNatequit (Quit: Computer has gone to sleep.)
03:09:37  <isaacs>wow, teh effect is NOT the same on smartos
03:10:03  <isaacs>i think the confounding thing here is that streams1 got slower on os x, that's all
03:10:07  <isaacs>and streams2 got just slightly faster.
03:10:21  <tjfontaine>what's the gap on smartos?
03:11:26  <isaacs>i'm running it
03:11:37  <isaacs>it looks like it's still around 5-7% slower with streams2
03:12:24  <isaacs>the testing is very tedious, becuase you have to wait 120 seconds at least between runs
03:12:34  <isaacs>and then the test itself takes 2 minutes
03:12:45  <tjfontaine>how fast does the vm reboot?
03:12:53  <isaacs>hahah, probably about that
03:12:56  <isaacs>maybe faster.
03:13:15  <isaacs>i bet there's probably some magic command to tell the kernel to release all the ephemeral ports
03:15:14  <isaacs>tjfontaine: https://docs.google.com/spreadsheet/ccc?key=0AganzoeqkiHddE9uRlA5WS0xSF9NTGN1QS0zZ3FRSWc#gid=2
03:15:23  <isaacs>the gap is pretty consistently 5% or so
03:17:02  <tjfontaine>still achievable, though I do wonder about the discrepency among platforms
03:17:28  <tjfontaine>it is frustrating that the gap isn't the same
03:19:19  * indexzeroquit (Quit: indexzero)
03:22:36  <isaacs>yea
03:22:45  <isaacs>i keep making changes that make os x faster, and have no effect on smartos
03:22:48  <isaacs>super frustrating.
03:23:17  <isaacs>oh!!!!!!!!
03:23:26  <isaacs>maybe it's a ia32/x86 discrepancy!!
03:23:49  <tjfontaine>now we're talking
03:29:44  * indexzerojoined
03:29:44  * indexzeroquit (Client Quit)
03:30:26  <isaacs>ugh.
03:30:30  <isaacs>it's slower with x64
03:30:47  <isaacs>like, significantly slower
03:31:10  <isaacs>but maybe it's even more significantly slower with x64 on streams1 :)
03:34:02  <tjfontaine>if you're lucky :)
03:36:04  <isaacs>nope
03:36:13  <isaacs>it's worse
03:36:16  <isaacs>>.<
03:36:33  <isaacs>i think today is done.
03:36:43  <isaacs>i'm gonna let this last benchmark run, and call it a week
03:36:47  <tjfontaine>sounds good
03:37:02  <tjfontaine>distance always makes it easier to reapproach
03:37:13  <isaacs>yeah
03:37:43  <isaacs>what's realy weird is that i saw better performance in one of the runs where i had dtrace profiling it 13 times per ms
03:37:52  <isaacs>which seems rather counterintuitive
03:38:57  <isaacs>yeah, still 5% slower on x64
03:38:58  <isaacs>about the same
03:39:16  <tjfontaine>I can't think of a plausible reason about why it would be faster under dtrace
03:40:13  <isaacs>yeah.
03:40:20  <isaacs>i can only attribute it to user error.
03:41:00  <isaacs>bnoordhuis gave me until 1/15 to make streams2 faster.
03:41:05  <isaacs>:)
03:41:12  <isaacs>so i've gotta finish this next week
03:41:19  <tjfontaine>heh, a few hours left then
03:41:33  <isaacs>well, few days
03:41:43  <isaacs>besides, we can just push back the release if necessary
03:41:59  <isaacs>it's already faster than v0.8
03:42:08  <tjfontaine>ya, there's nothing banging on your door to get .10 out is there?
03:42:52  <isaacs>oh, actually, no i'ts not faster than 0.8...
03:42:57  <isaacs>so why is streams1 slower??
03:42:59  * lohkeypart
03:43:02  <isaacs>that makes no sense. unless v8 got slower.
03:43:17  <isaacs>fuck, this is shitty.
03:43:26  <tjfontaine>well bnoordhuis did notice a couple other regressions
03:43:49  <isaacs>yeah
03:43:51  <isaacs>ugh.
03:44:20  <isaacs>ok, time to put this down for now, i guess
03:44:22  <tjfontaine>lots of moving pieces to this puzzle
03:44:33  <tjfontaine>v8, ev removal, streams2
03:44:47  <isaacs>HAHAH!! ok
03:44:55  <isaacs>so, v0.8's gap over streams2?
03:44:58  <isaacs>~5%
03:45:20  <isaacs>so this whole excitement and everythign? just that streams1+v8+master-uv has a regression
03:45:34  * isaacsfeeling sheepish
03:45:36  <tjfontaine>heh
03:45:44  <isaacs>oh well.
03:45:48  <isaacs>Monday.
03:45:53  <isaacs>have a good weekend :)
03:46:03  <tjfontaine>you too :)
03:46:08  * isaacs&
03:47:47  <isaacs>ircretary: tell piscisaureus Nevermind. The issue is that the streams1 js is triggering a V8 regression of some sort, apparently.
03:47:48  <ircretary>isaacs: I'll be sure to tell piscisaureus
04:10:20  * bradleymeckjoined
04:17:24  * brsonquit (Ping timeout: 248 seconds)
04:17:42  * lohkeyjoined
04:48:49  <Raynos>Whats the easiest way to run one test file in node core?
05:02:30  * TooTallNatejoined
05:07:08  * TooTallNatequit (Ping timeout: 252 seconds)
05:15:03  <bradleymeck>isaacs: execSync / spawnSync working thanks to streams2 being awesome, need to reread a ton looking for races
05:15:37  <bradleymeck>will cleanup and put on a branch tonight
05:23:43  * trevnorrisjoined
05:24:51  * sblomquit (Ping timeout: 245 seconds)
05:29:38  * bradleymeckquit (Quit: bradleymeck)
05:44:30  * brsonjoined
05:46:00  <Raynos>streams question.
05:46:29  <Raynos>Say you have `read.pipe(write)` and write returns true to apply back pressure and is then ended and goes into END state without emitting drain
05:46:42  <Raynos>then you do `read.pipe(write2)`
05:47:01  <Raynos>The read stream is not in the "still waiting on drain from ended stream" state and never flows into write2
05:47:20  <Raynos>is that a bug in pipe or a naughty userland write stream
05:50:14  * bradleymeckjoined
05:52:47  * brsonquit (Ping timeout: 252 seconds)
05:53:20  * brsonjoined
05:57:28  * jmar777quit (Remote host closed the connection)
05:58:04  * jmar777joined
05:59:51  <Raynos>I found the answer to said question in the source code, answer is yes pipe handles this but there was a seperate bug
06:02:30  * jmar777quit (Ping timeout: 264 seconds)
06:07:51  * mscdexjoined
06:08:18  * bradleymeckquit (Quit: bradleymeck)
06:08:33  <mscdex>ok, so i'm experiencing an oddity on Windows with fs.close() for node: i get an error passed to the callback that simply says: "Error: OK, close"
06:10:41  * brsonquit (Ping timeout: 248 seconds)
06:12:40  <mscdex>oops, i think i see what's wrong.. i was calling fs.close() many times in a row. but i think the error should be more descriptive :-)
06:12:40  * brsonjoined
06:15:38  * mscdexpart ("Leaving")
06:47:19  * mikealquit (Quit: Leaving.)
06:58:05  * mikealjoined
07:04:45  * lohkeyquit (Quit: lohkey)
07:16:14  * Raltjoined
07:23:29  * Raltquit (Remote host closed the connection)
07:24:37  * rendarjoined
07:56:10  * trevnorrisquit (Quit: Leaving)
08:03:54  * karupaneruraquit (Excess Flood)
08:06:31  * karupanerurajoined
08:08:32  * paddybyersjoined
08:47:00  * paddybyersquit (Ping timeout: 264 seconds)
08:58:35  * paddybyersjoined
08:58:53  * brsonquit (Read error: Connection reset by peer)
09:23:18  * loladiroquit (Quit: loladiro)
09:44:03  * felixgejoined
09:44:03  * felixgequit (Changing host)
09:44:03  * felixgejoined
09:44:37  * felixgequit (Client Quit)
09:44:46  * felixgejoined
09:44:46  * felixgequit (Changing host)
09:44:46  * felixgejoined
09:48:55  * felixgequit (Client Quit)
09:49:06  * felixgejoined
09:49:06  * felixgequit (Changing host)
09:49:07  * felixgejoined
09:49:20  * felixgequit (Client Quit)
09:49:32  * felixgejoined
09:49:32  * felixgequit (Changing host)
09:49:32  * felixgejoined
09:49:43  * felixgequit (Client Quit)
09:51:42  * stagasjoined
09:54:32  * felixgejoined
09:54:33  * felixgequit (Changing host)
09:54:33  * felixgejoined
09:59:30  * loladirojoined
10:07:25  * stagas_joined
10:09:08  * stagasquit (Ping timeout: 256 seconds)
10:09:20  * stagas_changed nick to stagas
10:17:20  * saghulquit (Ping timeout: 252 seconds)
10:18:07  * saghuljoined
10:20:52  * loladiroquit (Quit: loladiro)
10:58:15  * stagas_joined
10:59:02  * stagasquit (Ping timeout: 255 seconds)
10:59:09  * stagas_changed nick to stagas
11:00:56  * hzjoined
11:09:50  * AvianFluquit (Remote host closed the connection)
11:14:36  * `3rdEdenjoined
11:40:49  * paddybyersquit (Ping timeout: 248 seconds)
12:04:15  * `3rdEdenquit (Remote host closed the connection)
12:07:59  * felixgequit (Quit: felixge)
12:09:36  * TheJHjoined
12:18:13  * paddybyersjoined
12:34:46  * `3rdEdenjoined
12:43:13  * `3rdEdenquit (Ping timeout: 248 seconds)
12:51:12  * paddybyersquit (Ping timeout: 276 seconds)
13:08:05  * felixgejoined
13:08:05  * felixgequit (Changing host)
13:08:05  * felixgejoined
13:12:07  * hzquit (Disconnected by services)
13:12:11  * hzjoined
13:13:18  * TheJHquit (Ping timeout: 276 seconds)
13:34:42  * TheJHjoined
13:37:04  * paddybyersjoined
13:41:07  * V1joined
13:41:40  * V1changed nick to `3rdEden
14:06:13  * indutnyquit (Quit: Planned maintenance, back soon)
14:06:15  * dscapequit (Quit: Planned maintenance, back soon)
14:06:16  * Raynosquit (Quit: Planned maintenance, back soon)
14:06:16  * wavded_quit (Quit: Planned maintenance, back soon)
14:11:00  * geertjquit (Ping timeout: 248 seconds)
14:16:39  * felixgequit (Quit: felixge)
14:25:58  * TheJHquit (Read error: Operation timed out)
14:54:43  * indutnyjoined
14:56:02  * TheJHjoined
14:56:36  * dscapejoined
14:58:31  * Raynosjoined
15:01:05  * wavded_joined
15:03:18  * qmx|awaychanged nick to qmx
15:19:00  * stagas_joined
15:21:10  * stagasquit (Ping timeout: 240 seconds)
15:21:25  * stagas_changed nick to stagas
15:47:57  * bnoordhuisjoined
15:54:25  * felixgejoined
15:54:25  * felixgequit (Changing host)
15:54:25  * felixgejoined
16:16:39  * `3rdEdenquit (Remote host closed the connection)
16:27:35  <indutny>bnoordhuis: hey man
16:27:37  <indutny>how is it going?
16:27:43  <bnoordhuis>indutny: ho man
16:27:45  <bnoordhuis>good. you?
16:27:54  <indutny>good too
16:32:58  * loladirojoined
16:38:39  * loladiroquit (Quit: loladiro)
16:42:28  * qmxchanged nick to qmx|away
16:46:47  * bradleymeckjoined
16:47:20  * loladirojoined
16:50:07  * trevnorrisjoined
16:52:32  <isaacs>bradleymeck: whoa, really?
16:52:41  <bradleymeck>yea im looking for races
16:53:09  * mikealquit (Quit: Leaving.)
16:53:15  <bradleymeck>since streams2 buffers alread im just having it pick up the output and reusing most of the code with uv_run(different_loop)
16:53:26  <isaacs>bradleymeck: please get it into a pull req (marked as "not ready for inclusion, just for review" in the notes) so that piscisaureus et al can weigh in on approach etc.
16:53:27  <bradleymeck>have to change the exit behavior but thats mostly it
16:53:43  <bradleymeck>sure let me clean it up/get breakfast
16:53:55  <isaacs>Raynos: THe best way to run one test is ./node test/simple/test-whatever-blah.js
16:54:07  <isaacs>Raynos: if write() returns false, then that stream MUST eventually emit drain.
16:54:38  <isaacs>Raynos: if I stream.write(x) and then YOU stream.end(), well... that's undefined behavior
16:55:44  <isaacs>bradleymeck: surely
17:08:19  * trevnorrisquit (Ping timeout: 276 seconds)
17:18:25  * `3rdEdenjoined
17:26:40  * `3rdEdenquit (Ping timeout: 240 seconds)
17:28:50  * TheJHquit (Ping timeout: 255 seconds)
17:45:44  * AvianFlujoined
17:59:11  * mikealjoined
17:59:52  * loladiroquit (Quit: loladiro)
18:05:09  * loladirojoined
18:05:42  * TheJHjoined
18:17:27  * loladiroquit (Quit: loladiro)
18:17:53  * `3rdEdenjoined
18:22:44  * jmar777joined
18:32:16  * loladirojoined
18:34:39  * loladiroquit (Client Quit)
18:37:18  <isaacs>hm... i think the performance regression is in libuv, actually
18:37:21  <isaacs>could be, anyway
18:42:35  * jmar777quit (Remote host closed the connection)
18:43:07  * jmar777joined
18:48:06  * jmar777quit (Ping timeout: 264 seconds)
18:48:25  * bradleymeckquit (Quit: bradleymeck)
18:48:55  * stagas_joined
18:49:33  <indutny>isaacs: huh?
18:49:38  <indutny>isaacs: evidence?
18:50:06  <isaacs>indutny: oh, it seems that v0.8 is using v8 3.11.10.25 and my streams1 test branch is using 3.11.10.17
18:50:17  <indutny>yeah
18:50:24  <isaacs>indutny: but, master > streams1, and v0.8 is > both by a LOT
18:50:28  <isaacs>like 10-15%
18:50:30  * stagasquit (Ping timeout: 264 seconds)
18:50:34  * stagas_changed nick to stagas
18:51:11  <isaacs>will try with .25
18:51:39  <isaacs>btw, streams1 is usually master, with the lib/ folder from right before hte streams2 merge.
18:51:50  <isaacs>however, i noticed yesterday that it got a lot slower
18:52:03  <isaacs>at first, i thought stremas2 had gotten faster, and i got happy
18:52:13  <isaacs>then i realized i was wrong, and got sad
18:53:21  <indutny>isaacs: can you try updating v8 in v0.8?
18:53:39  <isaacs>indutny: that's another approach, of course.
18:53:52  <indutny>well, if you think v8 is the reason
18:53:55  <isaacs>indutny: right now, i'm trying v8 3.11.10.25 with master src and streams1 lib
18:53:56  <indutny>either degrade v8 version in master
18:53:59  <indutny>or upgrade in v0.8
18:54:10  <indutny>10.25?!
18:54:20  <indutny>aah
18:54:20  <isaacs>the v8 that v0.8 is using
18:54:21  <indutny>10.2
18:54:23  <indutny>ok
18:54:31  <indutny>got it
18:54:48  <isaacs>trying to figure out why node v0.8 is so much faster than streams1 all of the sudden
18:54:58  <isaacs>maybe it was before and i just didn't notice? i'm not sure
18:55:22  <indutny>it might be this
18:56:11  <isaacs>yeah, it's a bout the same
18:56:26  <isaacs>either the js in master pre-streams2 was slower, or the libuv in master is slower.
18:56:45  <indutny>I really doubt it's libuv changes
18:56:51  <indutny>because we have benchmarks in libuv
18:56:58  <indutny>and regression should be visible in them
18:57:09  <indutny>isaacs: on what OS are you benchmarking?
18:57:12  <isaacs>right, but what's the other factor?
18:57:13  <isaacs>os x
18:57:22  <isaacs>how do i run the benchmarks in libuv?
18:57:56  <isaacs>oh, make bench
18:57:56  <isaacs>derp
19:01:21  * TooTallNatejoined
19:01:31  * felixgequit (Read error: Connection reset by peer)
19:02:04  * felixgejoined
19:02:04  * felixgequit (Changing host)
19:02:04  * felixgejoined
19:02:07  <indutny>isaacs: hehe
19:02:15  <indutny>isaacs: either v8
19:02:19  <indutny>our bindings
19:02:21  <indutny>or js code
19:03:01  <isaacs>well, swapping out hte js code from streams1 to streams2, using the same libuv, bindings, and v8, seems to be faster at this point, on os x, and slightly slower on smartos
19:03:08  <isaacs>but node v0.8 blows both out of the water
19:04:07  <isaacs>using the v8 from v0.8, it's the same
19:05:02  * paddybyersquit (Ping timeout: 252 seconds)
19:11:22  * mikealquit (Quit: Leaving.)
19:11:58  * loladirojoined
19:14:01  * mikealjoined
19:14:11  <indutny>odd
19:15:03  <indutny>on both OSes?
19:15:51  <isaacs>indutny: so.. the benchmarks are confusing.
19:15:57  <isaacs>indutny: some things are faster, some are slower
19:16:11  <indutny>that always happens with JIT VMs
19:16:26  <isaacs>indutny: this is make bench in uv
19:16:30  <indutny>aah
19:16:32  <indutny>oow
19:19:28  <isaacs>indutny: https://gist.github.com/4520055
19:20:38  <indutny>hm...
19:20:45  <indutny>it seems that generally master is much faster than v0.8
19:20:52  <indutny>bnoordhuis: yt
19:20:53  <indutny>?
19:21:32  <isaacs>yeah.
19:21:35  <isaacs>i don't get it
19:23:50  <indutny>well
19:23:54  <indutny>it seems that libuv is faster
19:24:16  <indutny>so you say using v8 from v0.8 in master doesn't make benchmarks better?
19:24:29  <indutny>btw
19:24:32  <indutny>how are you benchmarking?
19:24:37  <indutny>cluster or single-process serevr?
19:24:38  <indutny>server*
19:24:43  <indutny>because there were some cluster changes
19:24:56  <indutny>and apparently it doesn't handle load well, right now
19:25:53  * Raltjoined
19:26:26  <isaacs>this is just a single process server
19:26:33  <isaacs>it's benchmark/http.sh
19:26:45  <isaacs>hold on, lemme run the streams1-new-v8 again. have to build it
19:26:49  <isaacs>and then i'll have another gist fo ryou
19:27:41  <indutny>ok
19:27:46  <indutny>then I suggest flamegraphing it! :)
19:27:48  <isaacs>hm... wait a second... my streams1 shows the node version as v0.9.4-pre
19:27:49  <isaacs>that's not right
19:27:56  <indutny>ahha!
19:28:10  <isaacs>oh, i think i moved back src/ so that it would work with the older v8
19:28:14  <isaacs>because we're doing the isolate stuff now
19:28:27  <isaacs>so that's not really a great test.
19:30:23  <isaacs>but youer' right, libuv is not the culprit, it would seem
19:30:40  <isaacs>building updated streams1 now
19:31:00  <isaacs>i have a movie date with my gf soon, though. so i can't spend all day doing this :)
19:31:31  * Raltquit (Remote host closed the connection)
19:31:56  <indutny>isaacs: turn off computer
19:31:57  <indutny>:)
19:32:02  <isaacs>haha
19:34:23  * bradleymeckjoined
19:35:43  <bradleymeck>isaacs: any sane way from node to invoke uv_read_start on the default allocator without needing a streamwrap (debugging something with large output), or can I move the OnRead/etc. with slab allocator to a public place to use from processwrap?
19:35:58  <bradleymeck>found some big bugs to work out before the PR
19:36:06  <isaacs>indutny: https://gist.github.com/4520094
19:36:31  <isaacs>indutny: so, on os x, master > streams1-with-same-deps-and-src
19:36:38  <isaacs>but v0.8 is still significantly faster than anything else.
19:36:39  <isaacs>so that sucs
19:36:45  <isaacs>bradleymeck: sorry, gotta run
19:36:49  <bradleymeck>sure
19:36:54  * isaacs&
19:42:44  * Raltjoined
19:43:37  * Raltquit (Remote host closed the connection)
19:45:54  * paddybyersjoined
20:00:09  <bradleymeck>anyone know decent way to just make an empty v8::Arguments
20:01:31  * `3rdEdenquit (Remote host closed the connection)
20:07:31  * sblomjoined
20:10:36  <bradleymeck>soo many edge cases with this damn thing
20:12:46  <MI6>joyent/node: Felix Böhm master * 9bce5e8 : repl: make built-in modules available by default Closes #3564. Closes #4 - http://git.io/fXvLwQ
20:15:07  <MI6>joyent/node: Nathan Rajlich master * bac537b : repl: fix lint - http://git.io/rsDoOA
20:15:30  * stagasquit (Ping timeout: 276 seconds)
20:21:29  * EhevuTovjoined
20:23:38  * Raltjoined
20:30:55  * qmx|awaychanged nick to qmx
20:35:43  * bradleymeckquit (Quit: bradleymeck)
20:36:10  * EhevuTovquit (Quit: This computer has gone to sleep)
21:07:19  * tomshredsjoined
21:11:57  * Raltquit (Remote host closed the connection)
21:18:59  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
21:25:44  * indexzerojoined
21:27:01  * Raltjoined
21:42:52  * indexzeroquit (Quit: indexzero)
21:51:24  * loladiroquit (Quit: loladiro)
21:55:42  * loladirojoined
21:57:49  * TooTallNatejoined
22:14:17  * felixgequit (Quit: felixge)
22:25:49  * sblomquit (Ping timeout: 256 seconds)
22:29:13  * rendarquit
22:31:36  * sblomjoined
22:33:27  * trevnorrisjoined
22:37:56  * Raltquit (Remote host closed the connection)
22:40:06  * qmxchanged nick to qmx|away
22:50:12  * indexzerojoined
22:54:34  * tomshredsquit (Quit: Leaving...)
22:55:27  * mikealquit (Quit: Leaving.)
22:57:33  * bradleymeckjoined
23:04:25  * bradleymeckquit (Quit: bradleymeck)
23:10:52  * mikealjoined
23:25:08  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
23:34:41  * trevnorrisquit (Quit: Leaving)
23:40:42  * EhevuTovjoined
23:45:27  * sblomquit (Ping timeout: 276 seconds)
23:45:43  * mikealquit (Quit: Leaving.)
23:49:06  * sblomjoined
23:55:45  * paddybyersquit (Ping timeout: 248 seconds)
23:58:02  * indexzeroquit (Quit: indexzero)