00:15:51  <bnoordhuis>isaacs, TooTallNate: https://github.com/bnoordhuis/node/commit/974299d <- review?
00:16:09  <CIA-108>node: isaacs v0.8 * re1fb7b7 / deps/uv/src/unix/proctitle.c : uv: Upgrade to b49d6f7 - http://git.io/lhbdNg
00:16:21  <CIA-108>node: isaacs v0.8.3-release * r14c8542 / (AUTHORS ChangeLog src/node_version.h): 2012.07.17, Version 0.8.3 (Stable) - http://git.io/X-cwLQ
00:16:59  <isaacs>bnoordhuis: Direct them towards using the domain module.
00:17:10  <isaacs>bnoordhuis: but still, yes, restarting the app is the best choice, even with domains
00:18:05  <bnoordhuis>isaacs: yeah, i intended to but then i realized that domains are not quite stable yet
00:19:06  <CIA-108>node: isaacs v0.8.3-release * ra6d730d / (AUTHORS ChangeLog src/node_version.h): 2012.07.17, Version 0.8.3 (Stable) - http://git.io/-o4Acw
00:20:14  <isaacs>pushed -RC1
00:20:27  <bnoordhuis>is it that time of the week again?
00:20:32  <isaacs>bnoordhuis: can you weigh in on the latest next-tick-now branch?
00:20:45  <isaacs>bnoordhuis: yeah, a bit past it actually. but there haven't been very many changes in v0.8.
00:20:50  <isaacs>it's actually, like, stable and stuff.
00:21:03  <bnoordhuis>isaacs: it already lgtm - the bug shigeki caught was pretty nifty though
00:21:19  <isaacs>yeah, it's the fix for that bug that i want to get a +1 on
00:21:25  <bnoordhuis>ah okay, one sec
00:21:51  <isaacs>https://github.com/joyent/node/commit/bb6033aa8d992493758a4211d82494ad7337e409
00:22:15  <isaacs>just read the description of the problem in teh commit.
00:22:21  <isaacs>that dude is some kind of edge-case-finding machine.
00:22:25  <isaacs>it's awesome
00:23:17  <bnoordhuis>yeah, he's awesome
00:23:38  <bnoordhuis>i wanted to snag him for c9 but alas
00:29:02  <bnoordhuis>isaacs: guess it looks okay to me
00:29:14  <bnoordhuis>mostly it makes my head hurt
00:29:51  <bnoordhuis>isaacs: btw, did you hand off that laptop to whats-his-name?
00:29:52  <isaacs>:D
00:29:55  <isaacs>bnoordhuis: yeah
00:30:11  <bnoordhuis>okay, cool - andrea morrow just mailed me about it
00:30:16  <isaacs>it's been recycled or donated to orphans or whatever happens to laptops
00:30:30  <bnoordhuis>or rather, that she hasn't had a confirmation yet
00:39:09  * EhevuTovquit (Quit: This computer has gone to sleep)
00:43:38  <bnoordhuis>full tree rebuilds are the bane of my existence...
00:49:54  * ericktquit (Read error: Operation timed out)
00:49:56  <bnoordhuis>isaacs: https://github.com/bnoordhuis/node/commit/43e9c83 ?
00:56:17  * dapquit (Quit: Leaving.)
01:12:27  * mikealjoined
01:21:38  * joeandaverdejoined
01:23:23  * joeandaverdequit (Client Quit)
01:28:17  <isaacs>bnoordhuis: TooTallNate found the issue I'd have.
01:28:20  <isaacs>fix that and lgtm
01:31:11  * mmaleckiquit (Quit: Reconnecting)
01:31:22  * mmaleckijoined
01:37:14  * c4milojoined
01:39:43  * EhevuTovjoined
01:41:55  * russell_hquit (*.net *.split)
01:42:18  <mmalecki>isaacs: hey, mind looking at one read issue real quick? kinda important and pressing: https://github.com/isaacs/read/pull/9
01:46:46  * russell_hjoined
01:48:18  <CIA-108>node: Ben Noordhuis v0.8 * re8af340 / doc/api/process.markdown : doc: advise *strongly* against uncaughtException - http://git.io/khageg
01:56:31  * felixgejoined
01:56:32  * felixgequit (Changing host)
01:56:32  * felixgejoined
01:56:35  * abraxasjoined
01:56:46  * ericktjoined
01:58:40  * felixgequit (Client Quit)
01:58:42  * TooTallNatequit (Quit: Computer has gone to sleep.)
02:01:25  * c4miloquit (Remote host closed the connection)
02:06:02  * bnoordhuisquit (Ping timeout: 252 seconds)
02:10:01  * EhevuTovquit (Quit: This computer has gone to sleep)
02:14:06  * EhevuTovjoined
02:14:22  <isaacs>mmalecki: sorry, i don't have much time at the moment to look at that
02:15:16  <isaacs>mmalecki: won't that break if you have an explicit num value?
02:15:31  <mmalecki>isaacs: okay, we'll float it for now then
02:15:59  <mmalecki>isaacs: well, how do you want to get an explicit num value there? you always get a char
02:16:36  <isaacs>mmalecki: read({num: 4})
02:16:40  <isaacs>mmalecki: then paste 5 chars
02:16:44  <isaacs>orjust type 4
02:17:02  <isaacs>oh, wait, i'm misreading that
02:28:26  * c4milojoined
02:34:36  * c4miloquit (Remote host closed the connection)
02:37:01  * beachdog_joined
02:37:50  * c4milojoined
02:40:08  * beachdogquit (Ping timeout: 252 seconds)
02:40:08  * beachdog_changed nick to beachdog
02:44:48  * c4miloquit (Remote host closed the connection)
02:48:06  * beachdogquit (Quit: beachdog)
03:11:07  * c4milojoined
03:38:35  * TooTallNatejoined
03:47:00  * c4miloquit (Remote host closed the connection)
04:35:55  * joeandaverdejoined
04:53:51  * TooTallNatequit (Quit: Computer has gone to sleep.)
04:59:57  * ericktquit (Quit: erickt)
05:15:52  * joeandaverdequit (Quit: joeandaverde)
05:40:25  * AvianFluchanged nick to AvianusAsleepus
05:48:33  * `3rdEdenjoined
05:52:31  * EhevuTovquit (Quit: This computer has gone to sleep)
05:58:20  * `3rdEdenquit (Quit: Leaving...)
06:13:33  * mikealquit (Quit: Leaving.)
06:28:19  * mikealjoined
06:30:32  * mikealquit (Client Quit)
06:32:05  * mikealjoined
06:41:20  * EhevuTovjoined
06:50:33  * stephankquit (Quit: *Poof!*)
07:02:59  * paddybyersjoined
07:09:22  * TooTallNatejoined
07:17:09  * paddybyersquit (Quit: paddybyers)
07:20:08  * TooTallNatequit (Ping timeout: 248 seconds)
07:35:29  * paddybyersjoined
07:36:22  * paddybyers_joined
07:39:52  * paddybyersquit (Ping timeout: 248 seconds)
07:39:52  * paddybyers_changed nick to paddybyers
08:20:37  * EhevuTovquit (Quit: This computer has gone to sleep)
08:49:51  * loladirojoined
09:05:12  * rendarjoined
10:02:21  * hzjoined
10:16:29  * V1joined
10:25:32  * V1quit (Quit: Leaving...)
11:01:07  * abraxasquit (Remote host closed the connection)
11:07:24  * hzquit (Ping timeout: 272 seconds)
11:08:12  * hzjoined
11:13:13  * bnoordhuisjoined
11:52:30  * beachdogjoined
11:57:44  * chobi_e_changed nick to chobi_e
11:58:13  <CIA-108>node: Ben Noordhuis v0.8 * rf6f2d42 / src/ev-emul.h : node: make ev-emul.h compile with -Wextra -Werror - http://git.io/XCYkBQ
12:25:20  <CIA-108>node: Ben Noordhuis v0.8 * rbeb7425 / README.md : doc: build requires GNU make 3.81+ - http://git.io/OJw4JA
12:27:15  * beachdogquit (Quit: beachdog)
12:48:59  * beachdogjoined
12:56:38  * hzquit (Read error: Connection reset by peer)
12:57:32  * hzjoined
13:00:41  <bnoordhuis>request for input - stop calling shutdown(SHUT_WR) on tcp sockets?
13:00:58  <bnoordhuis>the problem is that it can leave sockets hanging in the FIN_WAIT2 state
13:01:44  * beachdogquit (Remote host closed the connection)
13:01:54  * beachdogjoined
13:03:10  <indutny>bnoordhuis: why?
13:03:20  <bnoordhuis>indutny: why what?
13:03:31  <indutny>bnoordhuis: why sockets are hanging
13:04:04  <bnoordhuis>infidelity of the tcp protocol
13:04:16  <indutny>so client's ain't closing their ends
13:04:19  <bnoordhuis>yes
13:04:30  <bnoordhuis>the trouble is that the OS won't clean up the socket
13:04:32  <indutny>nice :)
13:04:40  <indutny>well, I think OS does what we're asking it for
13:05:00  <indutny>I completely do not mind
13:05:14  <indutny>but as I understand it may break some clients, right?
13:05:19  <bnoordhuis>yes, possibly
13:05:45  <indutny>my opinion: lets do it in 0.9 :)
13:05:56  <indutny>btw, how was we doing it before libuv has came?
13:06:44  <bnoordhuis>you mean in 0.4? it also did a shutdown()
13:06:55  <indutny>em...
13:07:05  <indutny>so why noone was complaining about that before?
13:07:14  <indutny>they just wasn't, right? :)
13:07:22  <bnoordhuis>probably because no one was running really large node installs
13:07:44  <indutny>well, how large
13:07:57  <bnoordhuis>something like 10-100K mobile clients
13:08:05  <indutny>well, I've runned such apps
13:08:14  <indutny>it was limited load
13:08:16  <indutny>but it was very high
13:08:26  <bnoordhuis>the mobile is relevant here because mobile users have the tendency to drop the connection without properly ending it
13:08:27  <indutny>like two hours of 1k-2k req/sec
13:08:35  <indutny>ah
13:08:47  <indutny>yeah, I think Mobile operators are just cutting this out
13:08:58  <indutny>or phones itself
13:09:04  <indutny>ok, nvm
13:09:30  <indutny>let me see how nginx is handling that
13:09:43  <indutny>I'm quite sure that noone has complained about it in nginx community
13:11:00  <bnoordhuis>seems it only calls shutdown() if the linger option is set?
13:12:12  <indutny>apperently yes
13:12:15  <indutny>apparently*
13:12:27  <indutny>still investigating
13:12:58  <bnoordhuis>apache does the same thing - only shutdown() when linger is enabled
13:13:10  <bnoordhuis>guess i know where igor got his inspiration from :)
13:13:26  <indutny>haha
13:13:37  <indutny>well, I'm quite sure he has looked at httpd :)
13:13:48  <indutny>handling all odd cases is really hard
13:14:07  <indutny>especially when building software that should fit existing complicated model
13:14:08  <indutny>like TCP
13:14:11  <bnoordhuis>yep
13:15:00  <indutny>so
13:15:04  <indutny>lets just remove it :)
13:15:10  <indutny>probably even from 0.8
13:15:14  <indutny>if isaacs do not mind
13:15:24  <bnoordhuis>not 0.8 but maybe in master
13:15:43  <bnoordhuis>guess we'll have to add a linger option as well
13:15:52  <bnoordhuis>ah, how oss projects converge
13:15:57  <indutny>hahahaha
13:23:41  <tjfontaine>I hear that mobile thing is going to fade out anyway, everyone is going to have 3 or 4 desktops in the house
13:26:18  <indutny>and in pockets
13:29:16  <bnoordhuis>cue the "is that a desktop pc in your pocket or are you just glad to see me?" jokes
13:30:02  <deoxxa>"hey baby, wanna play with my packard bell? ;)"
13:30:47  <mmalecki>wtf did I just read
13:31:18  <indutny>hahahaha
13:39:14  <indutny>oh, I just thought about following idea
13:39:27  <indutny>p2p network which allows sending/receiving json
13:39:38  <indutny>or multicasting json
13:39:55  <indutny>anyone knows such thing?
13:40:06  <indutny>ah, http://telehash.org/
13:41:10  * AvianusAsleepuschanged nick to AvianFlu
13:41:11  <deoxxa>i built a system like that, based on public/private key encryption and signing
13:41:28  <deoxxa>for entirely decentralised message distribution
13:41:45  <indutny>yeah, I mean like any app can use it
13:42:02  <indutny>if everyone will be connected in the network, where everyone is reachable
13:42:07  <deoxxa>yep
13:42:23  <indutny>you can anonymize traffic
13:42:26  <indutny>or just send files/messages
13:42:30  <deoxxa>bingo
13:43:09  <deoxxa>mine used a dht system and message digests for routing
13:43:48  <deoxxa>node ids were in the same keyspace as message digests, so it was pretty simple
13:44:16  <deoxxa>read up on the way bittorrent does it - it's where i drew most of my inspiration
13:46:37  <indutny>well, yes that's what I was talking about
13:49:05  * chobi_echanged nick to chobi_e_
13:50:34  * hzquit
14:06:50  * hzjoined
14:11:14  <tjfontaine>libuv/src/unix/darwin.c:68:11: warning: 'AbsoluteToNanoseconds' is deprecated: first deprecated in Mac OS X 10.8 [-Wdeprecated-declarations]
14:11:20  <tjfontaine>just so we're aware of that :)
14:12:46  <mmalecki>indutny: I wrote that kind of thing
14:12:52  <mmalecki>indutny: sec, I'll add you to the repo
14:13:06  <mmalecki>(not yet finished, obviously)
14:14:37  <indutny>oh noes
14:14:44  <indutny>10.8 depricated something
14:14:51  <indutny>I thought they have just made few UI tweaks
14:14:51  <indutny>:D
14:15:12  <indutny>mmalecki: interesting!
14:15:41  <mmalecki>indutny: isn't it :) ?
14:16:02  <mmalecki>me and AvianFlu have some mad science plans related to it!
14:17:48  <deoxxa>mmalecki: is this public?
14:18:00  <indutny>deoxxa: nope
14:18:08  <mmalecki>deoxxa: your github username?
14:18:08  <deoxxa>oh darn :<
14:18:11  <deoxxa>deoxxa
14:18:21  <mmalecki>I just need to finish up some things and I'll make it public :)
14:18:41  <mmalecki>deoxxa: there we go, added you
14:18:46  <deoxxa>awesome
14:19:02  <deoxxa>i'll be interested to see how you do things like message authority and trust
14:19:23  <mmalecki>well, I don't :)
14:19:34  <indutny>mmalecki: hm.. you ain't using udp
14:19:34  <deoxxa>ha
14:19:42  <indutny>how are you going to connect peers behind NAT ?
14:20:05  <tjfontaine>stun
14:20:08  <mmalecki>indutny: you only need one peer which can talk to both sides of the net
14:20:34  <mmalecki>basically, it just works as long as you're connected to at least one peer
14:20:59  <indutny>mmalecki: oh, so it's more like organize-my-cloud utility
14:21:05  <indutny>not connect-our-home-pcs
14:22:15  <mmalecki>indutny: yeah :)
14:22:30  <mmalecki>I mean, you can connect your home pcs too
14:22:58  <mmalecki>but using udp in the cloud, in cases when you should be paranoid about message delivery would be kinda silly
14:23:14  <mmalecki>unless I'd implement parts of tcp myself
14:23:24  <mmalecki>but I want to keep it simple
14:23:46  * AndreasMadsenjoined
14:24:33  * xaqjoined
14:50:54  * c4milojoined
14:52:57  * travis-cijoined
14:52:58  <travis-ci>[travis-ci] wwwouter/node#2 (master - dfce3ed : wouterm): The build is still failing.
14:52:58  <travis-ci>[travis-ci] Change view : https://github.com/wwwouter/node/compare/66ed63da7e93...dfce3ed51482
14:52:58  <travis-ci>[travis-ci] Build details : http://travis-ci.org/wwwouter/node/builds/1895594
14:52:58  * travis-cipart
14:59:46  * beachdog_joined
14:59:57  * beachdogquit (Read error: Connection reset by peer)
14:59:57  * beachdog_changed nick to beachdog
15:13:48  * beachdogquit (Ping timeout: 245 seconds)
15:14:59  * mikealquit (Quit: Leaving.)
15:17:46  * beachdogjoined
15:28:15  * beachdogquit (Ping timeout: 248 seconds)
15:31:50  * beachdogjoined
15:36:08  * xaqquit (Read error: Connection reset by peer)
15:36:38  * xaqjoined
15:40:33  * ericktjoined
15:42:39  * beachdogquit (Ping timeout: 276 seconds)
15:46:27  * beachdogjoined
15:53:37  * stephankjoined
16:05:40  * TooTallNatejoined
16:19:15  * mikealjoined
16:27:59  * loladiroquit (Quit: loladiro)
16:38:31  * dapjoined
16:40:58  * beachdogquit (Remote host closed the connection)
16:41:08  * beachdogjoined
16:45:07  * hzquit
16:51:27  * mikealquit (Quit: Leaving.)
16:51:53  * beachdog_joined
16:52:04  * beachdogquit (Read error: Connection reset by peer)
16:52:04  * beachdog_changed nick to beachdog
16:52:16  * dapquit (Ping timeout: 252 seconds)
16:53:02  * perezdquit (Read error: Connection reset by peer)
16:55:42  * arlolrajoined
16:57:59  * dapjoined
17:00:20  * EhevuTovjoined
17:01:16  <indutny>bnoordhuis: yt?
17:08:23  * beachdogpart
17:08:59  * beachdogjoined
17:11:58  * `3rdEdenjoined
17:27:42  * dapquit (Quit: Leaving.)
17:45:24  * theColejoined
17:54:00  * theColequit (Remote host closed the connection)
17:54:15  * theColejoined
18:08:07  <isaacs>hey guys
18:08:20  <isaacs>so, i'm gonna release 0.8.3, if there's no objections.
18:09:20  <indutny>hehe
18:09:20  <indutny>hi
18:10:24  <creationix>is it just me or is the node mailing list slowing down?
18:16:29  <indutny>what do you mean?
18:16:31  <indutny>less emails?
18:16:34  <indutny>bnoordhuis: ping
18:16:40  <creationix>indutny, right, less emails
18:17:19  <TooTallNate>isaacs: +1 since we got the process.title fix and build fixes
18:17:24  <isaacs>kewl
18:17:34  <indutny>creationix: yeah, I've noticed this too
18:17:41  <TooTallNate>basically bnoordhuis doing his thing as usual :)
18:18:17  <creationix>process.title fix?
18:18:28  <indutny>hey, that was my fix :D
18:18:34  <indutny>bnoordhuis has only added a comment to it
18:18:48  <indutny>I don't really care, though :p
18:19:01  <tjfontaine>I'm not convinced you don't care :P
18:19:37  * theColequit (Quit: theCole)
18:20:08  <indutny>tjfontaine: a question :)
18:20:13  <indutny>because you're around
18:20:21  <tjfontaine>I hope I don't disappoint
18:20:30  <indutny>suppose I've received a TCP connection with fd
18:20:44  <indutny>(actually, it's even a HTTP connection)
18:20:47  <indutny>and want to reply to it
18:21:02  <indutny>should I use send() and shutdown(fd, SHUT_WR) then?
18:21:11  <indutny>I'm asking because this doesn't work for me :D
18:21:46  <tjfontaine>haha, well I defer to other FOSS projects :P
18:22:11  <indutny>hahaha
18:22:12  <indutny>ok
18:22:19  <indutny>I'm just curious when I should start writing
18:23:29  <indutny>bnoordhuis: where are you when I need ya?! ;)
18:23:35  <tjfontaine>heh
18:31:04  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
18:34:07  * philips_quit (Excess Flood)
18:34:17  * philips_joined
18:34:41  <indutny>wow
18:34:51  <indutny>naive http server on C is slow
18:34:59  <indutny>really slow
18:35:04  <indutny>1800 req/sec
18:42:32  * loladirojoined
18:46:23  <isaacs>oh, whoops... looks like i spoke too soon
18:46:27  <isaacs>make pkg is segfaulting
18:46:40  * isaacsdistcleaned, trying again
18:53:17  <isaacs>hm, yep.
18:53:18  <isaacs>ln -fs out/Release/node node
18:53:18  <isaacs>out/Release/node tools/installer.js install
18:53:18  <isaacs>make[1]: *** [install] Segmentation fault: 11
18:53:20  <isaacs>make: *** [out/node-v0.8.3.pkg] Error 2
18:53:23  <isaacs>that's no good
18:53:54  <indutny>em...
18:54:02  <indutny>bad bad pkg
18:55:03  <isaacs>maybe some kind of x64/ia32 confusion?
18:56:09  <indutny>what `file ./out/Release/node` says?
19:00:10  <isaacs>indutny: $ file out/Release/node
19:00:11  <isaacs>out/Release/node: Mach-O executable i386
19:01:12  <indutny>well
19:01:32  <indutny>gdb out/Release/node
19:01:32  <indutny>r
19:01:35  <indutny>bt full
19:01:36  <indutny>?
19:01:54  <indutny>though I've 64bit node
19:02:01  <indutny>are you building it on i386 machine?
19:02:15  <isaacs>indutny: no, but it builds both ia386 and x64
19:02:24  <indutny>em..
19:02:30  <isaacs>then lipo's them
19:02:39  <indutny>well, run gdb then
19:02:42  <isaacs>er, ia32/i386
19:03:47  <isaacs>hold on, rebuilding node
19:04:35  <isaacs>i see, so it does this:
19:04:41  * arlolraquit (Quit: Linkinus - http://linkinus.com)
19:05:12  <isaacs>rm -rf out/deps out/Release
19:05:31  <isaacs>./configure --prefix=$(PKGDIR)/32/usr/local --without-snapshot --dest-cpu=ia32
19:05:35  <isaacs>$(MAKE) install
19:06:09  <isaacs>so even if i already have a out/Release/node, it's getting replaced with a new one
19:06:24  <isaacs>out/Release/node tools/installer.js install $(DESTDIR)
19:06:32  <isaacs>so that calls the newly built one
19:09:22  <TooTallNate>well nothing about that pkg process has changed right?
19:09:25  <TooTallNate>seems weird
19:10:03  <isaacs>TooTallNate: no, it must just be that it used to work, so we never noticed.
19:10:30  <isaacs>what it needs to do is make all first, then cp the out/Release/node somewhere safe, and then NODE=path/to/good/node make install
19:10:44  <isaacs>and use $(NODE) instead of out/Release/node everywhere
19:10:51  <isaacs>in teh meantime, i think i can kludge this manually
19:11:37  <TooTallNate>it's too bad gyp doesn't support fat binaries properly yet :\
19:11:57  <isaacs>yep.
19:11:59  <indutny>so
19:12:01  <tjfontaine>I think they're just waiting for it all to be 64bit only :)
19:12:04  <isaacs> install: all
19:12:04  <isaacs>- out/Release/node tools/installer.js install $(DESTDIR)
19:12:05  <isaacs>+ ./node tools/installer.js install $(DESTDIR)
19:12:17  <indutny>oooooh
19:12:18  <indutny>that thing
19:12:28  <isaacs>that's my temporary workaround
19:12:34  <isaacs>i just copied the real node to pwd
19:14:54  <creationix>lol, on semver.org: "How do I know when to release 1.0.0?
19:14:55  <creationix>If your software is being used in production, it should probably already be 1.0.0. If you have a stable API on which users have come to depend, you should be 1.0.0. If you're worrying a lot about backwards compatibility, you should probably already be 1.0.0."
19:15:54  <isaacs>ugh. that webpage is so preachy.
19:16:21  <isaacs>semver = slightly more humane version of a floating point number.
19:24:08  * ryahwaves
19:24:26  <ryah>TooTallNate: chrome only builds 32 bit binaries, riht?
19:24:35  <ryah>why do we even ship 64bit?
19:25:01  <TooTallNate>ryah: because most external libs (i.e. via "brew") will be 64-bit
19:25:11  <TooTallNate>another reason to push for static deps ;)
19:25:32  <tjfontaine>ryah: because when you compile on osx gyp detects it as 64bit if your system supports it
19:25:45  <ryah>oh right, because the entire unix infrastructure is fucked beyond help - i forgot
19:25:50  <isaacs>ryah: well, for starters, apparently the 32bit build doesn't actually run properly on a 64 bit os x
19:25:54  <isaacs>ryah: yes.
19:26:00  <isaacs>ryah: it's totally terrible.
19:26:03  * madsonjoined
19:26:10  <tjfontaine>meh, I'll avoid the troll bait
19:26:22  <isaacs>unix has failed.
19:26:24  <isaacs>it's time to admit it
19:26:35  <mmalecki>and windows didn't? come on.
19:26:43  <ryah>TooTallNate: any progress on the screencast?
19:26:44  <isaacs>mmalecki: i never said windows didnt
19:26:50  <mmalecki>okay, fine with me then
19:26:53  <tjfontaine>seriously library dependency (affectionately known as dll hell?) on windows is good?
19:27:01  <isaacs>mmalecki: there's plenty of fail to go around in computers
19:27:02  * xaq_joined
19:27:05  <tjfontaine>ok I'll change windows now
19:27:26  <mmalecki>tjfontaine: you're pull requesting windows?
19:27:44  <ryah>windows is also, obviously, fucked. even more so than unix :)
19:27:54  <ryah>if that's imaginable
19:27:58  * txdv_joined
19:28:28  * AlbireoX__joined
19:28:30  <tjfontaine>mmalecki: I can't imagine what would happen if I had to participate in that
19:28:56  <TooTallNate>ryah: just thoughts, nothing created yet. i got final cut pro though, so i'm excited to try it out
19:29:09  <TooTallNate>anyone know any good screen cap software?
19:29:26  <tjfontaine>quicktime or vlc :P
19:29:27  <TooTallNate>i need one for both os x and windows
19:29:51  <TooTallNate>ok, i'll poke around with them then
19:30:07  <TooTallNate>oh, probably quicktime pro right?
19:30:23  <TooTallNate>oh no… wow
19:30:30  <tjfontaine>nope, free for all
19:30:34  * pfox___joined
19:31:24  <isaacs>TooTallNate: quicktime is pretty nice
19:31:33  <isaacs>TooTallNate: not sure about windows, though
19:31:40  <TooTallNate>ill find something
19:32:10  <tjfontaine>I mean, vlc will work cross platform
19:32:15  <TooTallNate>hopefully this comp can handle vmware + windows + screen cap ;)
19:32:28  <tjfontaine>doesn't vmware have a tool for it?
19:32:42  <TooTallNate>vlc is great
19:32:52  <TooTallNate>idk, probably :D
19:32:53  <tjfontaine>ya workstation 5 has one, I dunno if it's in fusion
19:33:01  <TooTallNate>that sure would be convenient
19:33:36  <ryah>TooTallNate: i just sent you and mikeal and izs an email about this. mikeal knows a lot about this - we should get his advice
19:33:56  <ryah>good screencasts are so much better than bad screencasts
19:33:59  <tjfontaine>TooTallNate: I mean, you can always use quicktime and fusion full screen as a work around
19:34:04  <ryah>it's worth the time to think about it a bit
19:34:13  <TooTallNate>ryah: cool thanks, will do
19:34:31  <TooTallNate>I'm decent at editing fwiw. used final cut in high school a lot
19:34:36  <ryah>oh, good
19:34:49  * mjr_joined
19:35:03  * xaqquit (*.net *.split)
19:35:03  * AndreasMadsenquit (*.net *.split)
19:35:04  * txdvquit (*.net *.split)
19:35:04  * AlbireoX_quit (*.net *.split)
19:35:04  * pfox__quit (*.net *.split)
19:35:04  * ircretaryquit (*.net *.split)
19:35:04  * LOUDBOTquit (*.net *.split)
19:35:07  <ryah>to me editting is some magical process that i have never understood
19:37:27  <TooTallNate>apple makes it easy as usual
19:39:27  <creationix>I noticed that many linux distros come with video editing software in the default install now
19:39:43  <creationix>I think they want people to make screencasts to show off the desktop
19:39:52  <isaacs>TooTallNate: i'm not feeling apple making things easy atm..
19:40:06  <isaacs>0.8.3 is like 100% ready to go, looking great, just getting this #@$!# pkg to build properly
19:40:27  <isaacs>did we stop supporting ./configure --prefix=/x/y/z at some point?
19:40:35  <isaacs>it seems like make install still wants to go to /usr/local
19:40:38  <TooTallNate>have you gdb'd the segfault?
19:40:45  <TooTallNate>and no, no that I know of
19:41:43  <creationix>isaacs, I hope that's still supported, that's what nvm uses
19:41:58  * `3rdEdenjoined
19:43:18  <mmalecki>isaacs: no, that works for me with give
19:43:44  <bnoordhuis>indutny: pong
19:44:47  <isaacs>mmalecki: yeah, nave relies on that as well
19:44:56  <isaacs>i must be seeing a differentthing...
19:48:16  <ryah>isaacs: you have a new RC?
19:52:03  <indutny>bnoordhuis: nvm )
19:52:11  <indutny>I've built a http server without your help :D
19:52:18  <bnoordhuis>oh, good :)
19:52:36  <indutny>not in javascript
19:52:38  <indutny>in case if you're asking :D
19:53:11  <isaacs>ryah: http://nodejs.org/dist/v0.8.3/node-v0.8.3.tar.gz
19:53:22  <isaacs>ryah: but i'm wrestling with the pkg thing
19:53:35  <isaacs>otherwise, it works 100% everywhere
19:53:55  <isaacs>ryah: except your net crash thing, but that doesn't seem to be reproducible
19:54:02  <isaacs>at lesat, not easily
19:54:13  <ryah>yeah, seems like one of those spurious fails
19:57:02  <isaacs>ryah: don't get me wrong, it'd be great to track down
19:57:10  <isaacs>either the code is a little busted or the test is.
19:57:22  <isaacs>but 0.8.3 can go out withtout it
20:00:47  * `3rdEdenquit (Remote host closed the connection)
20:01:40  <ryah>yep
20:05:46  * paddybyers_joined
20:06:39  * paddybyersquit (Ping timeout: 248 seconds)
20:06:39  * paddybyers_changed nick to paddybyers
20:09:39  <isaacs>oh, crazy... this thing only segfaults when runing as sudo
20:13:10  <bnoordhuis>isaacs: what crashes? the installer or the node binary?
20:13:25  <isaacs>bnoordhuis: make, it looks like
20:13:39  * hzjoined
20:13:42  <isaacs>oh, no, it's the installer
20:14:02  <ryah>linux: [02:00|% 100|+ 437|- 1]: Done
20:14:09  * hzquit (Read error: Connection reset by peer)
20:14:14  <isaacs>yeah, it's just node
20:14:19  <isaacs>sudo node < -- crashes
20:14:24  <isaacs>ok, that's no good
20:14:34  <bnoordhuis>isaacs: sudo gdb --args node
20:14:39  <bnoordhuis>then r and bt full
20:15:23  <ryah>what does "bt full" do?
20:15:35  <tjfontaine>information overload usually :P
20:15:35  <ryah>as opposed to "bt"?
20:15:36  <bnoordhuis>prints a backtrace with locals
20:15:42  <ryah>ah
20:15:49  <tjfontaine>not as helpful in optimized builds though
20:15:53  <bnoordhuis>alas, no :/
20:16:49  <isaacs>ok, i'm confused: https://gist.github.com/3138586
20:17:36  <isaacs>it works in gdb, but not when running directly
20:17:54  <bnoordhuis>isaacs: heisenbugs are the best
20:17:55  <tjfontaine>sudo -s and when in there ulimit -c unlimited
20:17:59  <bnoordhuis>set ulimit -c unlimited
20:18:04  <bnoordhuis>ah, what tjfontaine said
20:18:35  <bnoordhuis>load the core dump with `gdb node core`
20:18:56  <pquerna>sudo sh -c "ulimit -c unlimited; node..."
20:19:15  * `3rdEdenjoined
20:20:11  <tjfontaine>the sudo environment barrier is sometimes hard to anticipate :)
20:20:59  <isaacs>hm.. it said that it dumped its core, but there's no core file.
20:21:48  <tjfontaine>is it your installed node? it could be beside the binary
20:21:50  <ryah>isaacs: in /cores ?
20:21:59  <tjfontaine>or there
20:22:07  <pquerna>kernel.core_pattern?
20:22:41  <isaacs>ah, ok, it's in /cores/core.$pid
20:23:01  <bnoordhuis>is /cores an os x thing?
20:23:05  <ryah>yes
20:23:19  <bnoordhuis>one lives and learns
20:23:39  <tjfontaine>pquerna: I wish most linux's shipped with saner core_pattern's
20:23:46  <pquerna>+1
20:24:09  <isaacs>ok... gdb says there's no stack, though
20:24:50  <tjfontaine>I'm seeing the same behavior, though my gdb is still loading symbols
20:24:55  <bnoordhuis>isaacs: thread apply bt full?
20:25:04  <ryah>isaacs: does this happen with node_g ?
20:25:39  <tjfontaine>#0 0x00007fff856e3ff9 in __findenv ()
20:25:40  <tjfontaine>in mine
20:25:49  <bnoordhuis>tjfontaine: is that the whole backtrace?
20:26:04  <tjfontaine>http://paste.debian.net/hidden/8d37a707/
20:26:21  <ryah>i bet it has to do with the setup for process.title = asdf
20:26:37  <ryah>don't we like write over the argv section or something?
20:26:41  <isaacs>hm, seems like dap's latest thing doesn't work on smartos
20:26:45  <isaacs>ryah: yeah.
20:27:02  <bnoordhuis>hmm, i wonder if that's because of fedor's proctitle patch
20:27:18  <isaacs>dtrace: failed to compile script src/v8ustack.d: line 194: failed to resolve V8DBG_OFF_FP_CONTEXT: Unknown variable name
20:27:30  <mmalecki>isaacs: OH. you're seeing that too?
20:27:34  <tjfontaine>ya mmalecki was complaing ...
20:27:36  <tjfontaine>:)
20:27:43  <isaacs>k, 0.8.3 is going out not-today :)
20:27:44  <ryah>rm dist/node-v0.8.3.targ.z
20:27:46  <ryah>:)
20:28:03  <bnoordhuis>isaacs: what happens when you revert libuv to 3726dee?
20:28:26  <bnoordhuis>s/revert/downgrade/
20:29:06  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
20:29:40  <mmalecki>isaacs: is that on smartos btw?
20:31:01  <isaacs>mmalecki: the dtrace thing? yes.
20:31:32  <mmalecki>isaacs: that'd match. not seeing it on 0.6, but seeing it on 0.8, with my modifications to node.d
20:31:37  <mmalecki>and few other places too
20:31:50  <mmalecki>but yeah, it fails for me
20:32:13  <mmalecki>but 0.8.2 was successfully compiled with dtrace on our balancers
20:32:16  <mmalecki>so not sure what's up
20:32:21  <mmalecki>sorry if I'm ebing chaotic
20:32:25  <bnoordhuis>[02:11|% 100|+ 438|- 0]: Done <- fwiw, make test on linux
20:32:31  <mmalecki>etoomuchcaffeine
20:33:00  <tjfontaine>bnoordhuis: btw, when you get a chance I did update that unref pullreq
20:33:56  <bnoordhuis>tjfontaine: yes, i saw. i'm waiting for bert to review it
20:34:04  <tjfontaine>ok
20:34:27  <isaacs>bnoordhuis: it crashes the same if i revert that
20:34:51  <bnoordhuis>isaacs: ah, that's odd
20:35:02  <isaacs>oh, wait, i just reverted the most recent "upgrade uv"
20:39:05  <isaacs>bnoordhuis: yeah, rolling uv back past all the proctitle stuff works
20:39:14  <isaacs>bnoordhuis: what is this even for, anyway?
20:39:32  <tjfontaine>on newer kernels the proctitle needs more nul's
20:39:36  <tjfontaine>linux kernels
20:39:39  <isaacs>ok..
20:39:43  <isaacs>but the first change
20:39:46  <tjfontaine>iirc the bug correctly
20:39:47  <isaacs>what was that for?
20:39:51  <isaacs>tjfontaine: that was the second bug
20:40:07  <isaacs>i mean, why did we even mess around with proctitle betwen 0.8.2 and 0.8.3?
20:42:20  <bnoordhuis>isaacs: because it didn't work on most platforms
20:42:38  <isaacs>bnoordhuis: it still doesn't work on os x or smartos, though
20:42:48  <bnoordhuis>it depends on the tool you use
20:42:49  <isaacs>from my pov, that's most platforms :)
20:42:56  * hzjoined
20:42:57  <isaacs>bnoordhuis: true, it works with ps on osx
20:43:44  <isaacs>but at the expense of not being able to run node as root. doesn't seem worth it.
20:43:45  * theColejoined
20:44:29  <bnoordhuis>running as root shouldn't matter
20:45:04  <bnoordhuis>it's kind of odd - that code originally came from a project of mine that's over 10 years old by now
20:45:41  <bnoordhuis>i assume someone would have complained about it by now
20:45:59  <tjfontaine>node_g doesn't bring any new information to the table either
20:46:15  <isaacs>bnoordhuis: reproducible on linux? or os x only?
20:47:34  <bnoordhuis>isaacs: probably os only. running as root doesn't crash here, i don't see memory corruption or anything either
20:47:58  <bnoordhuis>*os x
20:48:53  <bnoordhuis>i'll roll back libuv to the last known-good commit and fedor and i will look at it tomorrow
20:49:27  <isaacs>ok, cool
20:49:29  <isaacs>thanks
20:49:45  <isaacs>it looks like b40f813 is most likely the smartos-breaker
20:50:17  <isaacs>unless the blog post or the shebang tool broke it
20:50:49  <isaacs>it'd be nice if there was a wya to tell bisect not to bother with other branches
20:52:18  <indutny>isaacs: so are we good now?
20:52:33  <indutny>or are we still bad at socket test crash?
20:52:50  <isaacs>indutny: we're bad at smartos building and `sudo node` on osx segfaulting
20:52:57  <indutny>sudo nod
20:52:59  <indutny>hm...
20:53:06  * isaacssudo nods back
20:53:16  <indutny>:)
20:53:24  <isaacs>we can roll back the proctitle stuff.
20:53:34  <isaacs>but b40f813 is indeed the culprit on smartos
20:53:51  <indutny>heh, what's wrong with proctitle stuff?
20:55:15  <CIA-108>libuv: Ben Noordhuis v0.8 * r63c5f00 / (9 files in 4 dirs): Undo changes to uv_set_process_title(). - http://git.io/NDUYJg
20:55:44  <isaacs>indutny: not sure, but it makes node segfault when run as root
20:55:52  <CIA-108>libuv: Ben Noordhuis v0.8 * r69a6afe / (9 files in 4 dirs): unix: undo changes to uv_set_process_title() - http://git.io/VkspyA
20:56:09  <bnoordhuis>indutny: ^
20:56:21  <bnoordhuis>there's an example backtrace in the commit log
20:56:29  <indutny>hm...
20:56:42  <bnoordhuis>i can't reproduce it on linux
20:56:52  <indutny>that's on osx, right?
20:57:03  <indutny>I wonder how nginx works with this
20:57:32  <indutny>btw, it works fine for me on 10.6.8
20:57:35  * travis-cijoined
20:57:36  <travis-ci>[travis-ci] joyent/libuv#495 (v0.8 - 63c5f00 : Ben Noordhuis): The build is still failing.
20:57:36  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/b49d6f7c3042...63c5f00745d0
20:57:36  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/1898851
20:57:36  * travis-cipart
20:57:41  <indutny>should I do something special, or just `sudo node` ?
20:58:19  <bnoordhuis>isaacs: ?
20:58:41  <bnoordhuis>indutny: i tried both `sudo out/Release/node` and `sudo out/Release/node -e 'process.title = "BAM"'`
20:58:54  <isaacs>just `sudo node` causes the crash for me
20:58:55  * travis-cijoined
20:58:55  <travis-ci>[travis-ci] joyent/libuv#496 (v0.8 - 69a6afe : Ben Noordhuis): The build is still failing.
20:58:55  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/63c5f00745d0...69a6afea63fa
20:58:55  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/1898877
20:58:55  * travis-cipart
20:59:05  <indutny>em...
20:59:06  <indutny>both works
20:59:07  <tjfontaine>indutny: 10.6 ancient man :P
20:59:25  <indutny>tjfontaine: well, buy me a new mac please :)
20:59:34  <tjfontaine>hehe
20:59:46  <indutny>I really will enjoy it
20:59:50  <indutny>tjfontaine: where are you working at?
20:59:52  <isaacs>indutny: i'm on Darwin tum.local 11.4.0 Darwin Kernel Version 11.4.0: Mon Apr 9 19:32:15 PDT 2012; root:xnu-1699.26.8~1/RELEASE_X86_64 x86_64
20:59:53  <tjfontaine>I'll donate the $20 to get you lion :P
21:00:01  <indutny>tjfontaine: I can't install it
21:00:04  <indutny>due to corporate restrictions
21:00:23  <indutny>Apple has very strange license
21:00:30  <tjfontaine>for completeness sake I have: 12.0.0 Darwin Kernel Version 12.0.0: Sun Jun 24 23:00:16 PDT 2012; root:xnu-2050.7.9~1/RELEASE_X86_64 x86_64
21:00:33  <indutny>basically, if you want to update OS on corporate notebook - you should buy new
21:00:47  <indutny>Darwin indutny-osx.local 10.8.0 Darwin Kernel Version 10.8.0: Tue Jun 7 16:33:36 PDT 2011; root:xnu-1504.15.3~1/RELEASE_I386 i386
21:01:21  <indutny>can someone give me disassembly of function that causes sigsegv?
21:01:23  <tjfontaine>oh i386 as well, I wonder if that is related as well
21:01:42  <indutny>note that node is x64 anyway
21:06:17  <tjfontaine>indutny: http://paste.debian.net/hidden/276a5203/
21:06:25  <bnoordhuis>hmm, is github down? `git fetch` hangs
21:06:31  <bnoordhuis>on more than one machine btw
21:06:39  <tjfontaine>bnoordhuis: wfm
21:06:41  <bnoordhuis>oh, and as we speak...
21:06:46  <bnoordhuis>after hanging there for five minutes...
21:10:06  <indutny>tjfontaine: http://opensource.apple.com/source/OpenSSH/OpenSSH-7.1/openssh/bsd-setenv.c?txt
21:10:32  <indutny>so
21:10:36  <indutny>basically we're overwriting env
21:10:38  <indutny>and then using it
21:13:05  <indutny>odd
21:13:14  <indutny>can anyone give me an ssh access to segfaulting machine?
21:13:17  <indutny>:)
21:13:21  <indutny>I can send you public keys if that will help
21:13:26  <indutny>I think we can fix sudo problem
21:14:06  <tjfontaine>I can't today, but probably figure out something tomorrow
21:14:17  <indutny>tjfontaine: deal
21:14:31  <tjfontaine>sounds good
21:19:02  <indutny>tjfontaine: can you print environ variable
21:19:09  <indutny>at the time of segfault
21:19:33  <isaacs>indutny: one sec, i'm testing on another mac that is online all the time
21:19:58  <indutny>isaacs: cool
21:20:35  <TooTallNate>bnoordhuis: btw i can get you access to an OS X machine if you feel like looking into that /dev/null tihng
21:22:28  <indutny>isaacs: I suppose we've something wrong inside it
21:26:26  * xaq_quit (Remote host closed the connection)
21:27:36  <CIA-108>node: Ben Noordhuis v0.8 * r98b1ce9 / configure : build: fix mksnapshot crash on sunos - http://git.io/G8s93A
21:30:45  <isaacs>anyone remember the command to map port 2222 on some random server to localhost:22?
21:31:13  <bnoordhuis>isaacs: with ssh?
21:31:15  <indutny>ssh
21:31:20  <isaacs>yes, with ssh
21:31:29  <bnoordhuis>reverse or forward?
21:31:32  <indutny>let me google it for you
21:31:44  <indutny>ssh -f user@personal-server.com -L 2000:personal-server.com:25 -N
21:32:03  <bnoordhuis>or -R if you want to reverse-proxy it
21:33:57  <isaacs>yeah, that's not doing it
21:34:02  <indutny>em...
21:34:22  <isaacs>ok, i've got sshd running on port 2222 in debug mode
21:34:26  <isaacs>i cah ssh directly to it
21:34:31  <isaacs>but i want YOU to be able to ssh directly to it
21:35:06  <bnoordhuis>isaacs: -R to some machine fedor can log into
21:35:07  <indutny>just buy me a mac
21:35:08  <indutny>:D
21:35:19  <indutny>hahaha
21:36:01  <bnoordhuis>isaacs: ssh -R 2222:127.0.0.1:2222 user@machine-fedor-logs-into
21:36:36  <bnoordhuis>then fedor logs into said machine and does `ssh -p crazyisaac1980@2222 127.0.0.1`
21:36:53  <bnoordhuis>err, ssh -p 2222 crazyisaac1980@127.0.0.1
21:37:05  <indutny>oh my gosh
21:37:12  <bnoordhuis>easy as pie
21:37:33  <indutny>we should have a vpn for core team
21:38:40  <isaacs>so, i did 165.225.128.184
21:38:52  <isaacs>er ssh -R 2222:165.225.128.184:2222 165.225.128.184
21:39:25  <isaacs>and sshd is listening there
21:39:38  <bnoordhuis>isaacs: is 165.225.128.184 the target machine or the machine fedor should log into?
21:39:44  <isaacs>the machine to log into
21:39:55  <indutny>what's the target?
21:40:04  <bnoordhuis>isaacs: okay, you're almost there but use -R 2222:127.0.0.1:2222
21:40:32  <bnoordhuis>it means proxy port 2222 on the other machine to port 2222 on this machine
21:40:41  <isaacs>i thought there was a way i could just map port 2222 on the other machine to port 22 on this machine
21:40:58  <bnoordhuis>map, proxy, it's the same thing
21:41:08  <isaacs>so fedor does `ssh 165.225.128.184 -p 2222` and he's in
21:41:12  <bnoordhuis>yes
21:41:22  <indutny>Connection refused
21:41:22  <bnoordhuis>isaacs: "to port 22 on this machine" <- 22 or 2222?
21:41:35  <isaacs>bnoordhuis: doesn't matter, they've both got sshd on them :)
21:41:56  <bnoordhuis>indutny: sorry, `ssh 165.225.128.184`
21:42:05  <bnoordhuis>indutny: when logged in, `ssh -p 2222 127.0.0.1`
21:42:22  <indutny>ssh_exchange_identification: Connection closed by remote host
21:42:52  <bnoordhuis>indutny: when? first or second command?
21:42:56  <indutny>second
21:43:05  <bnoordhuis>isaacs: as what user should fedor log in?
21:43:09  <indutny>oh
21:43:10  <indutny>it works now
21:43:14  <indutny>em...
21:43:17  <indutny>what's the password?
21:43:18  <indutny>:)
21:43:24  <indutny>can you add my public key to that machine?
21:43:26  <bnoordhuis>hunter2
21:43:34  <isaacs>oh, jeez, adduser doesn't work
21:43:36  <indutny>bnoordhuis: --
21:43:38  <kohai>bnoordhuis has 19 beers
21:43:42  <bnoordhuis>aww :(
21:44:03  <isaacs>oh, right, OS X *Server*
21:44:06  <isaacs>@#!@$!@#$!@#
21:44:07  <indutny>isaacs: just add my public key
21:44:19  <indutny>isaacs: I've connected with key forwarding
21:44:39  <isaacs>anyway, nevermind, node isn't broken there
21:44:46  <isaacs>lol
21:44:51  <indutny>hahahaha
21:44:54  <indutny>so much pain
21:44:55  <indutny>and it works
21:44:56  <indutny>:D
21:45:04  <indutny>well, no rush
21:45:07  <indutny>lets do it tomorrow
21:45:12  <indutny>I'm going to sleep now anyway
21:45:15  <isaacs>k
21:45:32  <isaacs>woul'dve been nice to use a machine that will be awake when i'm asleep and you wake up :)
21:46:57  <indutny>hahaha
21:47:07  <indutny>all you need is VPN
21:47:10  <isaacs>yeah
21:47:18  <indutny>VPN VPN
21:47:22  <bnoordhuis>isaacs: if you set up a connection to an affected machine, i'll look at it
21:47:22  <indutny>VPN is all you need
21:47:46  <isaacs>hm... how do you create a user account on osx?
21:47:47  <indutny>bnoordhuis: it seems that we either not null terminating environ list
21:47:49  <indutny>after copying
21:47:49  <isaacs>this is so weird
21:47:56  <indutny>or not null terminating one of the strings
21:47:59  <bnoordhuis>isaacs: useradd?
21:48:03  <isaacs>yeah, not found
21:48:25  <bnoordhuis>http://osxdaily.com/2007/10/29/how-to-add-a-user-from-the-os-x-command-line-works-with-leopard/
21:48:39  <indutny>bnoordhuis: look at nginx's code
21:48:43  <indutny>bnoordhuis: I'm quite sure it works fine with sudo
21:50:15  <indutny>night everyoe
21:50:19  <indutny>closing my laptop
21:50:21  <indutny>ttyl
21:50:51  <isaacs>bnoordhuis: ok, you can log into that smartos drone, then -p 2222 goes to my laptop
21:51:34  * bnoordhuisopens yet another terminal
21:51:57  <isaacs>bnoordhuis: your key is there.
21:52:25  <bnoordhuis>isaacs: but apparently not on 165.225.128.184
21:52:55  <isaacs>bnoordhuis: log in as root there
21:53:06  <isaacs>bnoordhuis: or as node
21:54:33  <bnoordhuis>key exchanges on smartos machines always take so goddamn long
21:54:54  <isaacs>bnoordhuis: really? i find that they're usually snappy
21:55:05  <isaacs>could just be that i log in more frequently, so it's cached or something
21:55:10  <isaacs>it is making a network hop for it
21:55:24  <isaacs>a short one, but still
21:56:01  <bnoordhuis>isaacs: `ssh -p 2222 127.0.0.1` is asking me for a password
21:56:04  <bnoordhuis>and hunter2 isn't working
21:56:08  * rendarquit
21:56:21  <isaacs>bnoordhuis: did you sign into the smartos machine with -A?
21:56:23  <bnoordhuis>yes
21:56:30  * piscisaureus_joined
21:57:41  <isaacs>bnoordhuis: oh, wait, no do this: ssh -p 2222 bnoordhuis@127.0.0.1
21:58:15  <bnoordhuis>that works, thanks
21:58:28  <bnoordhuis>odd thing, bnoordhuis@127.0.0.1 works, bnoordhuis@localhost asks for a password
21:59:06  <piscisaureus_>hello
21:59:10  <piscisaureus_>how is it going?
21:59:34  <bnoordhuis>piscisaureus_: sup bertje?
21:59:37  <bnoordhuis>how's oscon?
21:59:45  <piscisaureus_>it's okay
21:59:52  <piscisaureus_>I talk to many people
21:59:54  <piscisaureus_>that's nice
22:00:22  * hzquit (Ping timeout: 272 seconds)
22:00:23  <bnoordhuis>do the people talk back?
22:00:23  <isaacs>bnoordhuis: because it's only bound on 127.0.0.1, not on localhost
22:00:29  <isaacs>bnoordhuis: localhost is actual localhost
22:00:50  <bnoordhuis>damnit, i lied - still asking for a pw
22:00:54  <isaacs>bnoordhuis: btw, you'll find that you're currently checking out node via git into ~bnoordhuis/node
22:01:35  <bnoordhuis>and hunter2 still doesn't work :/
22:01:42  <mmalecki>so, is this smartos thing fixed?
22:01:57  <bnoordhuis>mmalecki: mksnapshot crash?
22:02:08  <mmalecki>bnoordhuis: ustack something
22:02:31  <isaacs>bnoordhuis: ok, i hadn't enabled ssh for you
22:02:33  <isaacs>bnoordhuis: try now
22:02:44  <isaacs>bnoordhuis: hunter2 will never work on my laptop. that is way way way too insecure.
22:03:09  <bnoordhuis>isaacs: still asking for a pw
22:03:57  <isaacs>pm'd
22:04:20  * beachdogquit (Quit: beachdog)
22:04:29  <isaacs>bnoordhuis: ~bnoordhuis/node is building atm
22:05:14  <bnoordhuis>okay, cool
22:05:15  <isaacs>bnoordhuis: once it stops, i'll stop impersonating you
22:05:33  * chobi_e_changed nick to chobi_e
22:05:44  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
22:10:39  <CIA-108>node: Ben Noordhuis v0.8 * r5a5e128 / (common.gypi node.gyp): build: link with -rdynamic, not -Wl,--export-dynamic - http://git.io/mOr67w
22:11:13  <bnoordhuis>^ node builds again on solaris provided you configure with --without-dtrace
22:12:21  <isaacs>bnoordhuis: ok, all yours
22:12:35  * hzjoined
22:13:04  * hzquit (Client Quit)
22:14:00  <isaacs>bnoordhuis: can you build with less than -j4?
22:14:12  <bnoordhuis>sorry, yes :)
22:14:21  <bnoordhuis>-j 2 okay?
22:14:24  <isaacs>sure
22:14:31  <isaacs>(this is the part where ben sets my lap on fire remotely)
22:14:51  <bnoordhuis>awesome plot for a techno thriller
22:17:55  * c4miloquit (Remote host closed the connection)
22:17:55  <bnoordhuis>it is admittedly a complete mystery to me why gdb is so slow on macs
22:18:13  <bnoordhuis>esp. the 'Reading symbols for shared libraries' bit
22:28:16  * madsonquit (Remote host closed the connection)
22:39:49  * dapjoined
22:44:36  <mitsuhiko>is each uv_write a PSH on tcp?
22:44:56  <mitsuhiko>(eh, i guess it's called push)
22:45:29  <mitsuhiko>actually, i don't even know that's what's happening
22:45:42  <mitsuhiko>probably more like: is each ev_write a write + flush so to speak
22:46:10  <bnoordhuis>mitsuhiko: define flush?
22:47:11  <mitsuhiko>i'm ashamed to not know how this works internally, but i was generally assuming that something in the socket layer buffers stuff up
22:47:18  <mitsuhiko>which is not happening in libuv
22:47:40  <mitsuhiko>basically what it boils down to: do i need to buffer or is it okay to call uv_write with a few bytes each
22:48:00  <bnoordhuis>mitsuhiko: preferably buffer
22:48:12  <bnoordhuis>each uv_write more or less translates to a write() syscall
22:48:20  <bnoordhuis>syscalls are expensive are as small packets
22:48:56  <bnoordhuis>you can pass multiple uv_buf_t structs to uv_write btw and libuv will use writev() instead of write()
22:49:14  <bnoordhuis>s/are as/as are/
22:49:56  <mitsuhiko>indeed, i saw that but i elegantly ignored that so far ;)
22:50:04  <mitsuhiko>mostly because i actually never have more than one buffer object around
22:50:25  <mitsuhiko>i'm actually buffering into my own getto string builder thing
22:50:39  * piscisaureus_joined
22:51:03  <bnoordhuis>quiet guys, bert's back
22:51:31  <bnoordhuis>mitsuhiko: that works too - for high throughput, big buffers are best
22:52:52  <piscisaureus_>bnoordhuis: I am going ryah-ing
22:53:13  <piscisaureus_>bnoordhuis: drop by every now and then and just look from the sideline
22:53:30  <piscisaureus_>while you are sweating
22:53:38  <bnoordhuis>piscisaureus_: oh, i thought you meant "hang out with programmers"
22:54:14  <piscisaureus_>bnoordhuis: well, there are much more programmers in the physical "here" than there are in #libuv
22:55:49  <bnoordhuis>piscisaureus_: so what have you been up to? besides chatting up programmers
22:56:11  <piscisaureus_>bnoordhuis: we have to give a node training yesterday
22:56:33  <bnoordhuis>have or had?
22:56:45  <piscisaureus_>bnoordhuis: I sort of didn't want to go but apparently rik put up some text that said there would be a real node guy around
22:56:53  <bnoordhuis>hah
22:56:54  <piscisaureus_>bnoordhuis: so that could only be me
22:57:20  <piscisaureus_>bnoordhuis: had
22:57:23  <piscisaureus_>bnoordhuis: yesterday it was
22:57:31  <bnoordhuis>so what have you taught people?
22:57:47  <piscisaureus_>bnoordhuis: oscon is nice - the only thing I hate about it is the 10hr flight
22:58:06  <piscisaureus_>bnoordhuis: how to write node code
22:58:18  <piscisaureus_>bnoordhuis: I will give you details IRL if you really care,
22:58:24  <piscisaureus_>bnoordhuis: don't feel like typing it all out
22:58:37  <bnoordhuis>i will - i always like a good laugh
22:58:55  <bnoordhuis>couldn't c9 better have send someone who actually writes node code for a living?
22:59:07  <bnoordhuis>i mean, maintaining node is not quite the same thing as writing apps in it
22:59:23  <piscisaureus_>bnoordhuis: true
22:59:28  <piscisaureus_>bnoordhuis: I think that would've been better yeah
22:59:32  <bnoordhuis>i was pretty flabbergasted when people started asking me questions about express and step
22:59:44  <piscisaureus_>bnoordhuis: send sergi or jan would be good
23:00:01  <piscisaureus_>I am still trying to locate some people to troll here
23:00:12  <piscisaureus_>like Rasmus Lerdorf or something
23:00:19  <piscisaureus_>haha
23:00:21  <piscisaureus_>step
23:00:22  <piscisaureus_>oh my
23:00:32  <piscisaureus_>I know express somewhat
23:00:37  <piscisaureus_>but of course we're not teaching people that
23:01:43  <piscisaureus_>I am really confusted that there are people here that think perl is like state of the art
23:02:20  <deoxxa>piscisaureus_: i was being told about the vibrant ecosystem around java the other day
23:02:30  <deoxxa>piscisaureus_: you'll never stop being amazed
23:02:48  <piscisaureus_>indeed
23:03:00  <piscisaureus_>maybe I will get to that point
23:03:00  <deoxxa>oh and apparently c++ is dead
23:03:06  * philips_quit (Excess Flood)
23:03:07  <piscisaureus_>ehwat?
23:03:17  <deoxxa>just another gem from the same guy
23:03:19  <bnoordhuis>it does smell somewhat
23:03:38  <deoxxa>i honestly didn't know what to say to him to try to correct his broken view
23:04:03  <piscisaureus_>I just agree with them
23:04:16  * philips_joined
23:04:17  <piscisaureus_>and then I start raving about how awesome ant is
23:04:17  <deoxxa>i changed the subject
23:04:20  <piscisaureus_>and springsource
23:04:22  <deoxxa>haha
23:04:35  <bnoordhuis>piscisaureus_: maven > ant
23:04:45  <piscisaureus_>oh, yeah
23:04:51  <bnoordhuis>(both are awful actually)
23:04:52  * dapquit (Quit: Leaving.)
23:04:52  <piscisaureus_>I always forget the name :-)
23:05:05  <piscisaureus_>I have to run again
23:05:20  <piscisaureus_>I am trying to get into a vim vs emacs discussion after that
23:05:26  <deoxxa>we usually end up with a couple of people like that at the node meet here in melbourne
23:05:37  <deoxxa>best bet is to try to line them up with each other
23:05:50  <piscisaureus_>:-)
23:05:53  <piscisaureus_>sorry, gtg
23:05:59  <deoxxa>o/
23:06:28  <bnoordhuis>hf, bertje
23:08:08  <bnoordhuis>gdb and debugging in general on macs is truly, truly awful...
23:10:36  * piscisaureus_quit (Ping timeout: 260 seconds)
23:25:36  * c4milojoined
23:42:04  * beachdogjoined
23:52:17  * paddybyersquit (Quit: paddybyers)
23:58:19  * loladiroquit (Quit: loladiro)
23:58:46  <mitsuhiko>bnoordhuis: especially the ancient version of gdb on mac
23:59:02  <mitsuhiko>(thanks gnu people for gplv3 :P)