00:13:28  * AvianFluquit (Ping timeout: 240 seconds)
00:14:08  * AvianFlujoined
00:15:46  * mralephquit (Quit: Leaving.)
00:19:32  * isaacsjoined
00:52:35  <mmalecki>bnoordhuis: hey, is assignment to struct member in C slower than using local variable?
00:55:57  <bnoordhuis>mmalecki: the answer to that question is definitely maybe
00:56:49  <mmalecki>haha, ok. can struct members be optimized and moved to registers?
00:56:59  <bnoordhuis>mmalecki: yes
00:57:20  <mmalecki>ok, thanks
01:50:52  * theColequit (Quit: theCole)
02:18:15  * indexzeroquit (Quit: indexzero)
02:24:50  * brsonquit (Quit: leaving)
02:49:50  <mmalecki>bnoordhuis: https://github.com/mmalecki/http-parser/commit/b7136a73ead07adaa2e87d633eba9c011b1958ad
02:57:49  <mmalecki>bnoordhuis: in reference to http://groups.google.com/group/nodejs-dev/browse_thread/thread/11ee75aa49f2f1ab
03:26:30  * mmalecki_joined
03:26:58  * sj26quit (Ping timeout: 248 seconds)
03:28:06  * sj26joined
03:28:18  * bnoordhuisquit (Read error: Operation timed out)
03:28:32  * mmaleckiquit (Ping timeout: 248 seconds)
03:29:59  * mmalecki_quit (Client Quit)
03:30:02  * mmaleckijoined
03:53:17  * isaacs_mobilejoined
03:58:15  * indexzerojoined
03:59:17  * dshaw_quit (Quit: Leaving.)
04:02:06  * orlandovftwjoined
04:03:36  * isaacs_mobilequit (Remote host closed the connection)
04:25:43  * orlandovftwquit (Ping timeout: 245 seconds)
04:39:07  * mmaleckichanged nick to mmalecki[zzz]
04:58:03  * isaacsquit (Remote host closed the connection)
05:02:07  * ryahjoined
05:03:30  * orlandovftwjoined
05:03:37  * orlandovftwquit (Client Quit)
05:03:51  * orlandovftwjoined
05:20:19  * mikealjoined
05:32:07  * mikealquit (Quit: Leaving.)
05:47:48  * mikealjoined
06:16:17  * mikealquit (Quit: Leaving.)
06:16:48  * ryanjoined
06:17:14  * ryanchanged nick to Guest72448
06:19:59  * Guest72448quit (Client Quit)
06:20:16  * mikealjoined
06:21:13  * startup_joined
06:32:41  * startup_quit (Quit: Page closed)
06:40:10  * mikealquit (Quit: Leaving.)
07:03:52  * indexzeroquit (Quit: indexzero)
07:06:32  * paddybyersjoined
07:46:48  * stephankjoined
08:22:34  * benviequit
08:54:42  * bentkus_joined
08:58:28  * bentkusquit (Ping timeout: 265 seconds)
09:12:09  * bentkus_quit (Ping timeout: 252 seconds)
09:14:27  * bentkusjoined
09:41:27  <bbenvie>hmmmm node seems to work fine on Windows 8 except npm decides to install modules to C:\node_modules
09:43:05  <bbenvie>whatever anyone says about Windows, it sure is consistent about working for 2 decades after a given version
09:44:07  <bbenvie>I still don't forgive it for not supporting xterm-256 though
09:56:25  * mmalecki[zzz]quit (*.net *.split)
09:56:25  * DrPizzaquit (*.net *.split)
09:59:56  * mmalecki[zzz]joined
09:59:56  * DrPizzajoined
10:05:16  * sh1mmerquit (Ping timeout: 260 seconds)
10:12:28  * ryahquit (Quit: Lost terminal)
10:13:52  * indutny_sleepingchanged nick to indutny
10:49:19  * orlandovftwquit (Quit: leaving)
10:49:35  * AndreasMadsenjoined
11:46:38  * bbenviequit
12:13:03  * einarosjoined
12:26:27  * benviejoined
12:42:41  * bnoordhuisjoined
12:54:43  <CIA-110>node: Ben Noordhuis upgrade-gyp * r62ac960 / (121 files in 33 dirs): gyp: upgrade to r1197 - http://git.io/0V_Ciw
13:07:33  * mmalecki[zzz]changed nick to mmalecki
13:08:51  <mmalecki>good morning!
13:09:30  * travis-cijoined
13:09:30  <travis-ci>[travis-ci] joyent/node#406 (upgrade-gyp - 62ac960 : Ben Noordhuis): The build failed.
13:09:30  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/commit/62ac960
13:09:30  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/670874
13:09:30  * travis-cipart
13:09:48  <indutny>mmalecki: morning
13:10:16  <bnoordhuis>indutny: afternoon
13:12:30  <indutny>bnoordhuis: evening
13:17:41  <indutny>bnoordhuis: how are you doing, Ben?
13:18:06  <bnoordhuis>indutny: i'm doing well, fedor. how about you? :)
13:19:01  <indutny>bnoordhuis: me too
13:19:03  <indutny>:)
13:19:17  <indutny>doing some front-end stuff atm
13:19:52  <bnoordhuis>indutny: css > c++?
13:20:05  <indutny>bnoordhuis: haha
13:20:47  <mmalecki>indutny: btw, call in 40?
13:21:00  <indutny>mmalecki: yes
13:23:41  <mmalecki>ok, so I can do some reddit!
13:23:50  <mmalecki>^ me being productive
13:32:18  * bbbbjoined
13:38:49  <mmalecki>bnoordhuis: so, have you tried `give` yet :) ?
13:39:05  <bnoordhuis>mmalecki: not yet
13:39:18  * bbbbquit (Ping timeout: 276 seconds)
13:40:29  <mmalecki>bnoordhuis: ok. ping me if you find any problems :)
13:42:13  <mmalecki>fail, it's my most watched repo now
13:55:34  * piscisaureus_joined
13:57:02  <piscisaureus_>bnoordhuis: lgtm
13:57:23  <bnoordhuis>piscisaureus_: cool
13:57:41  <piscisaureus_>bnoordhuis: did you also add the mapping for ERROR_ALREADY_EXISTS?
13:57:46  <bnoordhuis>piscisaureus_: no
13:57:47  <CIA-110>libuv: Ben Noordhuis v0.6 * rc1cea70 / src/win/error.c : windows: map WSAETIMEDOUT to UV_ETIMEDOUT - http://git.io/2KrEEA
13:57:49  <piscisaureus_>hmm
13:58:13  <bnoordhuis>i have to leave something for you, right?
13:58:27  <piscisaureus_>bnoordhuis: no, I but I see the mapping is already there
13:59:31  * travis-cijoined
13:59:32  <travis-ci>[travis-ci] joyent/libuv#91 (v0.6 - c1cea70 : Ben Noordhuis): The build is still failing.
13:59:32  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/f9be43a...c1cea70
13:59:32  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/671083
13:59:32  * travis-cipart
14:02:46  * wankdankerjoined
14:08:23  * bentkusquit (Ping timeout: 244 seconds)
14:10:12  <bnoordhuis>is it just me or do github labels no longer work?
14:10:13  * paddybyersquit (Quit: paddybyers)
14:11:23  <bnoordhuis>it seems i can no longer tag issues... the label box is there, it folds out when i click on it but actually selecting a label no longer works
14:11:52  <piscisaureus_>bnoordhuis: why does uv_fs_s have `int errorno` instead of `uv_err_t error` ?
14:11:56  <piscisaureus_>bnoordhuis: :-(
14:12:19  <bnoordhuis>piscisaureus_: i don't know, i didn't design it
14:12:23  <piscisaureus_>bnoordhuis: how did this fall through review :-( :-(
14:12:32  <mmalecki>bnoordhuis: works for me
14:12:45  <bnoordhuis>hrm, let me try it in chrome
14:12:47  <mmalecki>indutny: also, get on skype, would you?
14:12:53  * bnoordhuistries to remember his github password
14:13:05  <piscisaureus_>bnoordhuis: g1rlp0w3r
14:13:05  <mmalecki>bnoordhuis: copy a cookie :>
14:13:13  <bnoordhuis>piscisaureus_: hey, thanks!
14:14:16  <bnoordhuis>using the same 5 digit password for every website is so awesome!
14:14:45  <bnoordhuis>hmm, doesn't work in chrome either...
14:14:46  <mmalecki>argh, another CLA
14:14:57  <piscisaureus_>12345
14:15:06  <bnoordhuis>piscisaureus_: https://github.com/joyent/node/issues/2744 <- can you tag this 'repl'?
14:15:27  <piscisaureus_>bnoordhuis: no, broken
14:15:32  <piscisaureus_>let me try internet explorer :-)
14:17:32  <piscisaureus_>IE9 is actually pretty decent
14:17:40  <piscisaureus_>but not gh seems to broken
14:17:44  <piscisaureus_>let me tweet
14:21:11  <bnoordhuis>mmalecki: you said it works for you?
14:24:14  <mmalecki>bnoordhuis: yes, on haibu repo
14:24:25  <piscisaureus_>bnoordhuis: it works again
14:24:34  <piscisaureus_>bnoordhuis: I tagged the issue
14:24:52  <bnoordhuis>ah, good
14:25:19  <piscisaureus_>bnoordhuis: but it's still broken. Setting one tag accidentally worked, but I can't add another one
14:27:03  <mmalecki>bnoordhuis: I'll fill the CLA later today, I have no idea what to put into this corporate field
14:27:35  <mmalecki>fucking legal stuff
14:27:42  <piscisaureus_>mmalecki: which CLA?
14:27:48  <mmalecki>piscisaureus_: parser cla
14:28:03  <piscisaureus_>mmalecki: for http parser?
14:28:09  <mmalecki>piscisaureus_: yes
14:28:12  <piscisaureus_>huh
14:28:15  <piscisaureus_>I didnt know that
14:28:20  <bnoordhuis>you do know
14:28:34  <piscisaureus_>bnoordhuis: there's a special one for http parser?
14:28:40  <bnoordhuis>yes
14:28:52  <piscisaureus_>bnoordhuis: I have no commit bits for http parser, and also never did a patch for it
14:29:10  <piscisaureus_>How were I supposed to know?
14:29:28  <bnoordhuis>hah, typo: you do *now*
14:29:49  <bnoordhuis>that k shouldn't have been in there :)
14:30:04  <piscisaureus_>bnoordhuis: do you have commit bits for it? or does ryan still gate keep that?
14:30:11  <bnoordhuis>i'm a committer
14:30:27  <bnoordhuis>probably *the* committer right now because no one else seems to work on it
14:30:33  <bnoordhuis>now for something completely different, i'll be in the office wednesday
14:30:42  <piscisaureus_>bnoordhuis: ok, nice
14:31:08  <piscisaureus_>bnoordhuis: now put that in the IDE chat :-)
14:31:24  <mmalecki>btw, you should put http_parser on travis, bnoordhuis
14:33:01  <bnoordhuis>mmalecki: i think i'm a contributor, not an admin
14:33:31  <mmalecki>bnoordhuis: lame. ask somebody, I'll send you the right .travis.yml later
14:36:05  * bentkusjoined
14:37:16  <piscisaureus_>hmm
14:37:30  <piscisaureus_>error handling in win/fs.c is a mess
14:37:54  <mmalecki><obvious comment about windows being a mess>
14:40:32  <AndreasMadsen>any new on this: https://github.com/joyent/node/issues/2598 - it's a real pain
14:41:10  <mmalecki>AndreasMadsen: ++, would be cool to see this fixed
14:41:10  <kohai>AndreasMadsen has 7 beers
14:43:45  <AndreasMadsen>My case is I'm calculating some 30 dimensional matrixes for an experminal neural network, but it takes like forever - so I would do it in a worker but sending the calculation instructions blocks :(
15:19:44  <piscisaureus_>mmalecki: indutny: with forever, is it possible to use --trace-gc?
15:22:07  <indutny>piscisaureus_: should be
15:22:57  <piscisaureus_>indutny: bt how?
15:24:48  <mmalecki>piscisaureus_: sec, let me try that, I think it should be with one hack
15:26:16  <mmalecki>piscisaureus_: do have some file that outputs something under --trace-gc?
15:26:25  <mmalecki>how do I know it's running?
15:26:34  <piscisaureus_>mmalecki: it writes stuff to stderr
15:35:50  <mmalecki>piscisaureus_: hm. it's complicated
15:36:19  <mmalecki>piscisaureus_: we check for existance of file if command is == node
15:36:33  <mmalecki>it's kind of lame
15:39:16  <mmalecki>piscisaureus_: sec, I see it
15:39:25  <mmalecki>piscisaureus_: I can disable it. I'll gist you a script to do it
15:43:02  <mmalecki>also, I'll try to get rid of node-fork, this thing is a real pain
15:43:03  <piscisaureus_>mmalecki: thanks
15:43:10  <piscisaureus_>we are now stepping through forever to get it done :-)
15:43:27  <mmalecki>piscisaureus_: oh! <3 <3 <3
15:44:45  <piscisaureus_>hmm. daemon.start() is crashing
15:49:44  * bbbbjoined
15:54:45  <mmalecki>piscisaureus_: ok, so, like, this is a bad, bad hack :(
15:55:32  <mmalecki>https://gist.github.com/1817817
15:55:38  <mmalecki>bbbb: indutny ^
15:55:45  <mmalecki>this shouldn't look like this guys
15:56:21  <indutny>mmalecki: heh
15:56:22  <indutny>not good
15:56:34  <mmalecki>piscisaureus_: crashing, how?
16:06:01  * indexzerojoined
16:06:26  <mmalecki>indutny: https://gist.github.com/1817817 makes puppies cry :(
16:06:31  <mmalecki>* indexzero ^
16:10:12  <bnoordhuis>https://github.com/joyent/node/issues/2747 <- one more reason to add uv_external_fd
16:12:10  <piscisaureus_>mmalecki: thanks
16:12:14  <piscisaureus_>mmalecki: makes sense
16:12:22  <piscisaureus_>mmalecki: and it made the puppy cry
16:12:40  <piscisaureus_>bnoordhuis: nooo
16:12:54  <piscisaureus_>bnoordhuis: that guy should just use net.Socket
16:13:29  <bnoordhuis>piscisaureus_: he's reading from arbitrary fds
16:13:31  <mmalecki>piscisaureus_: https://github.com/nodejitsu/forever/blob/master/lib/forever/monitor.js#L215 <- main reason
16:13:59  <piscisaureus_>bnoordhuis: so? you should be able to create a net.Socket from an arbitrary fd
16:14:17  <bnoordhuis>yeah, it might work, a pty is a stream
16:14:26  <mmalecki>piscisaureus_: I'll tweak this if, actually
16:14:47  <piscisaureus_>bnoordhuis: I will *not* have read(), write(), IoWatcher, setsockopt(), socket() bindings
16:15:20  <bnoordhuis>piscisaureus_: that's phase 2 :)
16:15:30  <piscisaureus_>bnoordhuis: no fucking way no
16:15:36  <bnoordhuis>but it would be convenient for add-ons to integrate with the event loop
16:16:13  <piscisaureus_>come to the office and we will disuss it
16:16:36  <bnoordhuis>the problem with net.Socket() is that it watches the fd for reading and writing
16:16:57  <piscisaureus_>bnoordhuis: so?
16:17:05  <piscisaureus_>bnoordhuis: it works fine with stdin right?
16:17:28  <piscisaureus_>bnoordhuis: but yeah we can tweak that a little to also work on half duplex streams
16:17:49  <bnoordhuis>hm, i don't know why it works for stdin
16:18:11  <piscisaureus_>bnoordhuis: well, because we never try to write to it
16:18:42  <piscisaureus_>bnoordhuis: and you probably can just add the fd to the write set
16:18:50  <piscisaureus_>but it'll never flag
16:19:01  * isaacsjoined
16:19:51  <bnoordhuis>piscisaureus_: node -e 'var fd = require("net").Socket(0); fd.write("BAM")'
16:19:58  <bnoordhuis>i would expect that to start busy looping but it doesn't
16:20:19  <piscisaureus_>bnoordhuis: why do you expect it to busy loop?
16:20:40  <bnoordhuis>oh wait, fds 0-2 don't have the O_NONBLOCK flag set
16:20:47  <piscisaureus_>bnoordhuis: yes they do
16:21:09  <piscisaureus_>bnoordhuis: fd 0 definitely have
16:21:11  <piscisaureus_>*has
16:21:24  <bnoordhuis>node never calls fcntl or ioctl on fd 0
16:22:09  <piscisaureus_>bnoordhuis: that would be a bug I thing
16:22:21  <piscisaureus_>bnoordhuis: node 0.4 definitely did
16:22:26  <bnoordhuis>yes
16:22:28  <piscisaureus_>and it also used net.stream for stdio
16:22:36  <piscisaureus_>(unless the thing was a file)
16:22:51  <piscisaureus_>bnoordhuis: https://github.com/joyent/libuv/blob/master/src/unix/tty.c#L40
16:23:31  <bnoordhuis>well, it seems that's never executed right now :)
16:24:02  <bnoordhuis>ah, not until you open process.stdin
16:24:07  <piscisaureus_>hweh
16:24:19  <piscisaureus_>yes, that's initialized lazily
16:25:01  <piscisaureus_>bnoordhuis: but anyway, we should have some fine grained control over this
16:25:12  <piscisaureus_>bnoordhuis: but not by supplying all these bindings.
16:25:35  <piscisaureus_>bnoordhuis: instead we just need to make sure that net.Socket and net.Server do the right thing(tm)
16:25:40  <isaacs>hooray, stdiod
16:26:08  <piscisaureus_>isaacs: what?
16:26:13  <bnoordhuis>piscisaureus_: agreed
16:26:40  <bnoordhuis>also, epoll_ctl(3, EPOLL_CTL_ADD, 4, {EPOLLIN, {u32=4, u64=4294967300}}) = 0
16:26:50  <piscisaureus_>bnoordhuis: there's more use cases for this. For example, creationix was trying to open /dev/js0 which is pollable
16:26:57  <bnoordhuis>that's why writing to stdin doesn't busy loop
16:26:59  <piscisaureus_>so we should not force people to use the thread pool there
16:27:17  <bnoordhuis>yes, agreed
16:27:19  <piscisaureus_>bnoordhuis: what does that mean?
16:27:42  <bnoordhuis>EPOLLIN = report readable
16:27:55  <isaacs>er, "hooray, stdio"
16:27:56  <bnoordhuis>EPOLLOUT (which is not set) = report writable
16:28:16  <piscisaureus_>bnoordhuis: oh, right. we probably just add it to the write set when we get EAGAIN
16:28:21  <piscisaureus_>bnoordhuis: which makes sense :-)
16:28:37  <piscisaureus_>bnoordhuis: otherwise a normal socket would also busyloop if we were not trying to send data
16:28:43  <bnoordhuis>yes, exactly
16:29:04  <bnoordhuis>so with ptys for example you need to update the read/write flags
16:29:15  <bnoordhuis>because a pty is nearly always writable
16:29:30  <bnoordhuis>hence the inevitable return of IOWatcher :)
16:29:35  <piscisaureus_>bnoordhuis: nope.
16:29:41  <bnoordhuis>or something like it
16:29:52  <piscisaureus_>bnoordhuis: for what do you need the IOWatcher now?
16:30:10  <bnoordhuis>ptys, systemd integration, that kind of thing
16:30:28  <piscisaureus_>bnoordhuis: so what can you do with IOWatcher that you can't do with net.Socket?
16:30:33  <piscisaureus_>or net.Server
16:30:37  <bnoordhuis>change the read/write flags
16:30:41  <creationix>pollable files are neat
16:30:52  <bnoordhuis>net.Socket() always watches for read *and* write events
16:31:02  <piscisaureus_>bnoordhuis: not true.
16:31:24  <bnoordhuis>piscisaureus_: https://github.com/joyent/libuv/blob/4e1f2b1/src/unix/pipe.c#L176
16:31:37  <piscisaureus_>bnoordhuis: the read event is watched if the socket is not paused. the write event is juggled based on the write queue size.
16:31:49  <piscisaureus_>bnoordhuis: that's something else
16:32:07  <piscisaureus_>bnoordhuis: that is for tracking whether we got EOF and/or sent FIN
16:32:26  <bnoordhuis>piscisaureus_: nuh uh, that's what PipeWrap::Open() calls
16:33:16  <bnoordhuis>now it's true that uv__stream_open() doesn't actually start the read/write watchers
16:33:30  <piscisaureus_>bnoordhuis: yes and it also doesn't look right :-)
16:33:46  <piscisaureus_>bnoordhuis: it should not start off with both event watched.
16:34:14  <piscisaureus_>bnoordhuis: oh wait - yes, that's it
16:34:23  <piscisaureus_>bnoordhuis: it just creates the watcher but they are not started
16:34:37  <piscisaureus_>bnoordhuis: because they are only started when appropriate
16:35:20  <piscisaureus_>bnoordhuis: btw - we recently added support for half-duplex pipes on windows. Unix should also be able to open data streams, like devices, ptys etc.
16:35:51  <piscisaureus_>half duplex data streams, that is
16:37:03  <mmalecki>piscisaureus_: https://github.com/nodejitsu/forever/commit/b093bfcfe2daee2f4c3ad7ca949531872ef40477
16:37:23  <piscisaureus_>mmalecki: awesome. thanks.
16:37:45  <mmalecki>piscisaureus_: my pleasure :). hit me up if you find any problems
16:37:52  <piscisaureus_>mmalecki: the latest forever, does that work with 0.4 still?
16:38:26  <mmalecki>piscisaureus_: I think it should, yeah. I'll check, se
16:41:17  <mmalecki>piscisaureus_: yes
16:42:00  <mmalecki>piscisaureus_: so, you guys wanted to write some daemonization in node core, right?
16:44:51  <AndreasMadsen>Oh, that would be really nice to have in core
16:45:01  <CIA-110>node: Karl Skomski master * r09ccbef / configure : build: detect host_arch better - http://git.io/GPTNVA
16:45:08  <mmalecki>piscisaureus_: I find Ruby API quite nice: Process.daemon http://www.ruby-doc.org/core-1.9.3/Process.html#method-c-daemon
16:49:10  * dapjoined
16:57:22  <mmalecki>bnoordhuis: will you deal with http_parser upgrade in node?
16:57:50  <isaacs>why do we need daemonization in core?
16:59:11  <isaacs>(but yes, the ruby api for this is quite nice)
16:59:18  <mmalecki>isaacs: because it's common when writing sysadmin tools/services, also it differs between platforms (so it's a thing libuv could handle)
16:59:51  * travis-cijoined
16:59:51  <travis-ci>[travis-ci] joyent/node#407 (master - 09ccbef : Karl Skomski): The build is still failing.
16:59:51  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/3f43b1c...09ccbef
16:59:51  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/672010
16:59:51  * travis-cipart
16:59:57  <isaacs>we'd probably never be able to do fork(function () { doStuff }) like ruby can
17:00:06  <piscisaureus_>and we don't want to :-)
17:00:10  <isaacs>yeah
17:00:17  <isaacs>how could that even work in ruby, anyway?
17:00:19  <piscisaureus_>although that would be nice
17:00:27  <piscisaureus_>well, under the hood
17:00:36  <isaacs>wouldn't it leak refs to stuff across the boundary?
17:00:53  <piscisaureus_>well you end up having 2 processes
17:00:56  <isaacs>or seem like it *should*, and then not, and be confusing and magically and terrible?
17:01:05  <piscisaureus_>*shrug*
17:01:07  <piscisaureus_>anyway
17:01:07  <mmalecki>isaacs: ruby has a weird threading model, only one thread running atm, iirc
17:01:08  * isaacsjust remembered ruby's design goals...
17:01:17  <mmalecki>lol
17:01:20  <isaacs>magically terrible is like item #2
17:01:23  <isaacs>anyway
17:01:38  <piscisaureus_>we can add daemonize bindings, windows services are very different
17:01:59  <isaacs>detach and daemonize sound nice, but a) not for v0.6, and b) lower priority than a bunch of other things we need for 0.8
17:02:03  <piscisaureus_>yes
17:02:10  <piscisaureus_>agreed
17:02:12  <isaacs>put in the post-0.8 pile.
17:02:16  <mmalecki>does it mean 'pull requests welcome'?
17:02:28  <isaacs>mmalecki: it means, pull requests welcome, but don't expect them to be merged in the next month or so.
17:02:41  <piscisaureus_>isaacs: well, pull requests are welcome, but review does not have a high prio
17:02:45  <isaacs>right
17:02:48  <mmalecki>isaacs: I can deal with it
17:02:49  <bnoordhuis>mmalecki: pull requests are always welcome
17:02:51  <piscisaureus_>er, mmalecki
17:03:01  <bnoordhuis>i haven't reached my 'issues closed' quota this week yet
17:03:13  <isaacs>mmalecki: even bad ideas are always welcome. the only reason i'd say no is to save your time doing something that's probably dumb :);
17:03:25  <isaacs>but more activity is generally a good thing
17:04:15  <mmalecki>isaacs: ok. I'll try writing this thing for unix when I find some cycles
17:15:50  * TooTallNatejoined
17:23:31  * philipsquit (Ping timeout: 260 seconds)
17:23:52  * philipsjoined
17:26:11  * orlandovftwjoined
17:26:58  <piscisaureus_>bnoordhuis: is there a way to look at a file and tell whether it is a socket, a fifo, a device, or an ordinary file?
17:27:22  <mmalecki>piscisaureus_: on unix it's stat
17:27:36  <piscisaureus_>can that tell a fifo from a socket?
17:28:24  <mmalecki>piscisaureus_: yes
17:28:47  * sh1mmerjoined
17:28:49  <piscisaureus_>we should just make net.createConnection do the right thing then
17:28:51  <mmalecki>piscisaureus_: are you on unix machine? I'm pretty sure there are macros in kernel files
17:28:57  <piscisaureus_>I am not
17:29:29  <isaacs>indutny: you'd left me a link to the v8 issue, but i seem to have misplaced it
17:29:56  <mmalecki>piscisaureus_: http://lxr.oss.org.cn/ident?i=S_ISFIFO
17:31:01  <mmalecki>piscisaureus_: S_ISFIFO(stat->st_mode) should work, I think
17:31:44  <piscisaureus_>we should make net.Socket a base class for net.Connection and net.Server
17:31:44  <isaacs>indutny: oh, found it. http://codereview.chromium.org/9382033/
17:51:26  * mikealjoined
17:51:34  * igorzijoined
17:58:00  <isaacs>bnoordhuis: thanks for merging v0.6 yesterday :D
17:58:18  <bnoordhuis>my pleasure :)
17:59:15  <indutny>isaacs: heya
17:59:23  <isaacs>indutny: yo
17:59:34  <indutny>isaacs: pinging Erik Corry right now
17:59:38  <isaacs>indutny: great
17:59:40  <isaacs>thanks
17:59:58  <indutny>isaacs: but I think we should try landing this in master
18:00:00  <indutny>and then in v0.6
18:00:22  <indutny>isaacs: it's quite simple
18:00:52  <indutny>isaacs: erik seems to be non-responsive for a few daays
18:02:03  <CIA-110>libuv: Bert Belder reviewme * r205522b / (include/uv-private/uv-win.h src/win/error.c src/win/fs.c): Clean up error handling in win/fs.c, add some error mappings - http://git.io/YcbUJw
18:02:17  <piscisaureus_>^-- igorzi: can you review that? I made error handling a little simpler, we now always translate the windows error to an uv error immediately.
18:03:45  * travis-cijoined
18:03:45  <travis-ci>[travis-ci] joyent/libuv#92 (reviewme - 205522b : Bert Belder): The build failed.
18:03:45  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/commit/205522b
18:03:45  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/672381
18:03:45  * travis-cipart
18:10:10  <igorzi>piscisaureus_: reviewing..
18:12:09  <CIA-110>node: isaacs master * r5041bc1 / (deps/v8/SConstruct deps/v8/build/common.gypi):
18:12:09  <CIA-110>node: Patches floating on V8
18:12:09  <CIA-110>node: Only SConstruct and build/common.gypi at this point. All
18:12:09  <CIA-110>node: others are accepted upstream, which greatly simplifies things. - http://git.io/MGuIzw
18:12:12  <CIA-110>node: isaacs master * r68a0c56 / (112 files in 13 dirs): Upgrade V8 to 3.9.5 - http://git.io/IQ01Bg
18:12:42  <isaacs>indutny: just updated v8 ^
18:12:55  <isaacs>indutny: feel free to land in master.
18:13:58  <TooTallNate>bnoordhuis: re: gyp-upgrade, we should wait until this lands: http://codereview.chromium.org/9382044/
18:14:08  <TooTallNate>so that we can build 64-bit on OS X
18:14:08  <CIA-110>libuv: Bert Belder reviewme * rb367c2a / (src/win/error.c test/test-fs.c): win: add ERROR_FILENAME_EXCED_RANGE mapping, fix fs_file_nametoolong test - http://git.io/t1uzCA
18:14:29  * mikealquit (Quit: Leaving.)
18:15:22  * orlandovftwquit (Ping timeout: 248 seconds)
18:15:28  <bnoordhuis>TooTallNate: noted
18:15:49  * travis-cijoined
18:15:49  <travis-ci>[travis-ci] joyent/libuv#93 (reviewme - b367c2a : Bert Belder): The build is still failing.
18:15:49  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/205522b...b367c2a
18:15:49  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/672430
18:15:49  * travis-cipart
18:20:19  <igorzi>piscisaureus_: shouldn't it also check for FILE_ATTRIBUTE_DIRECTORY before returning UV_EISDIR?
18:27:48  * travis-cijoined
18:27:48  <travis-ci>[travis-ci] joyent/node#408 (master - 5041bc1 : isaacs): The build is still failing.
18:27:48  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/09ccbef...5041bc1
18:27:48  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/672423
18:27:48  * travis-cipart
18:28:10  <piscisaureus_>igorzi: no
18:28:40  <piscisaureus_>igorzi: the problem is this: people do _open("c:\\windows", _O_RDWR)
18:28:54  <piscisaureus_>igorzi: that path already exists but it is a dir
18:29:07  <piscisaureus_>so we should do what unix does and say EISDIR
18:29:12  <piscisaureus_>but windows says ERROR_FILE_EXISTS
18:29:29  * mikealjoined
18:32:13  <igorzi>open("some_file", _O_CREAT | _O_TRUNC | _O_EXCL)? won't CreateFile also return ERROR_FILE_EXISTS?
18:32:24  <igorzi>piscisaureus_: --^
18:32:25  <kohai>piscisaureus_ has 4 beers
18:32:42  <igorzi>piscisaureus_ ++
18:32:43  <kohai>piscisaureus_ has 5 beers
18:32:44  <igorzi>:)
18:32:44  <piscisaureus_>igorzi: I'm not sure
18:32:50  <piscisaureus_>:-)
18:33:19  <piscisaureus_>igorzi: O_EXCL normally triggers EEXIST when the target already exists
18:33:45  <piscisaureus_>the question is what _O_TRUNC | _O_RDONLY would do
18:34:06  <igorzi>piscisaureus_: i'm just asking if it's ok to just rely on ERROR_FILE_EXISTS + (flags & _O_CREAT) to return UV_EISDIR. should we also make sure that path is actually a directory before returning UV_EISDIR?
18:34:28  <piscisaureus_>yeah you could have a point there
18:34:32  <piscisaureus_>what else could break it?
18:38:35  <igorzi>piscisaureus_: i think as long as you check for FILE_ATTRIBUTE_DIRECTORY, you should be good
18:38:58  <piscisaureus_>igorzi: from `man 2 open`
18:38:59  <piscisaureus_>EISDIR
18:38:59  <piscisaureus_> pathname refers to a directory and the access requested involved writing (that is, O_WRONLY or O_RDWR is set).
18:39:15  <piscisaureus_>EISDIR
18:39:16  <piscisaureus_> pathname refers to a directory and the access requested involved writing (that is, O_WRONLY or O_RDWR is set).
18:39:34  <piscisaureus_>EEXIST
18:39:34  <piscisaureus_> pathname already exists and O_CREAT and O_EXCL were used.
18:39:56  <igorzi>piscisaureus_: right, so EISDIR should only be returned if path is a directory
18:40:02  <piscisaureus_>yes
18:40:09  <piscisaureus_>but this, of course, is racy
18:40:33  <igorzi>true
18:40:45  <piscisaureus_>What else could trigger ERROR_FILE_EXISTS when O_CREAT is set?
18:41:44  <piscisaureus_>igorzi: I couldn't think of anything?
18:41:50  <piscisaureus_>s/?/!/
18:42:26  <igorzi>MSDN for CreateFile says ERROR_FILE_EXISTS is returned when CREATE_NEW is used
18:44:15  <piscisaureus_>igorzi: yes - that is not a lie
18:44:28  <piscisaureus_>igorzi: however it also does it when trying to open a directory for writing :-)
18:44:32  <piscisaureus_>igorzi: https://github.com/joyent/node/issues/2706
18:44:41  <igorzi>piscisaureus_: yep, that i figured :)
18:46:09  <piscisaureus_>igorzi: so probably I should check for error == ERROR_FILE_EXISTS && (flags & _O_CREAT) && !(flags & _O_EXCL)
18:46:19  * CoverSlidejoined
18:48:36  <igorzi>piscisaureus_: yeah, that should work
18:50:23  <piscisaureus_>igorzi: I assume you can sign this off too? https://github.com/joyent/libuv/commit/b367c2a276ab052200bc75a87ba37122f48b6d47
18:51:41  <bnoordhuis>piscisaureus_: _O_TRUNC | _O_RDONLY <- not valid, EINVAL
18:51:54  <piscisaureus_>right
18:52:30  <piscisaureus_>bnoordhuis: ok thanks, I was just exploring an edge case that doesn't happen :-)
18:54:05  <igorzi>piscisaureus_: yep, lgtm
18:54:21  <bnoordhuis>piscisaureus_: if you'd ask marc lehmann, he'd whip out his copy of the posix specification and declare with a nasal voice that the results are unspecified
18:54:39  <bnoordhuis>but in the real world you get EINVAL
18:54:50  <bnoordhuis>or maybe EACCES
18:55:31  <bnoordhuis>piscisaureus_: oh, maybe i spoke too soon - The (undefined) effect of O_RDONLY | O_TRUNC varies among implementations. On many systems the file is actually truncated.
18:56:08  <bnoordhuis>i wonder what systems those are
18:58:58  * mjr_joined
18:59:12  <piscisaureus_>bnoordhuis: well, on windows it doesn't work :_)
18:59:55  <piscisaureus_>CreateFile can't do it because NtCreateFile's api doesn't support it
19:01:30  <isaacs>piscisaureus_: should i update libuv in node master?
19:01:48  <piscisaureus_>isaacs: not yet
19:01:57  <isaacs>k
19:02:04  <isaacs>any idea when?
19:02:28  <CIA-110>libuv: Bert Belder v0.6 * rbc8b990 / (src/win/error.c test/test-fs.c): win: add ERROR_FILENAME_EXCED_RANGE mapping, fix fs_file_nametoolong test - http://git.io/Tmfrtw
19:02:33  <CIA-110>libuv: Bert Belder v0.6 * r6c80bf3 / (include/uv-private/uv-win.h src/win/error.c src/win/fs.c): Clean up error handling in win/fs.c, add some error mappings - http://git.io/fLo2uw
19:02:37  <piscisaureus_>isaacs: We have to merge v0.6 to master in libuv
19:02:43  <piscisaureus_>let me do that first
19:03:04  <isaacs>looks like bnoordhuis did yesterday
19:03:24  <piscisaureus_>isaacs: well - let me do it again. should be relatively easy then
19:03:51  <isaacs>k
19:04:12  * travis-cijoined
19:04:12  <travis-ci>[travis-ci] joyent/libuv#94 (v0.6 - bc8b990 : Bert Belder): The build is still failing.
19:04:12  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/c1cea70...bc8b990
19:04:12  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/672645
19:04:12  * travis-cipart
19:05:30  * orlandovftwjoined
19:05:41  * orlandovftwquit (Client Quit)
19:05:56  * orlandovftwjoined
19:06:02  <CIA-110>libuv: Bert Belder master * r1d942e2 / (4 files in 3 dirs): Merge branch 'v0.6' - http://git.io/5rORlA
19:06:04  <CIA-110>libuv: Ben Noordhuis master * rc1cea70 / src/win/error.c : windows: map WSAETIMEDOUT to UV_ETIMEDOUT - http://git.io/2KrEEA
19:06:05  <CIA-110>libuv: Bert Belder master * rbc8b990 / (src/win/error.c test/test-fs.c): win: add ERROR_FILENAME_EXCED_RANGE mapping, fix fs_file_nametoolong test - http://git.io/Tmfrtw
19:06:06  <CIA-110>libuv: Bert Belder master * r6c80bf3 / (include/uv-private/uv-win.h src/win/error.c src/win/fs.c): Clean up error handling in win/fs.c, add some error mappings - http://git.io/fLo2uw
19:06:15  <piscisaureus_>Hmm. [% 100|+ 118|- 7]: Done.
19:06:32  <piscisaureus_>Time to fix some test failures
19:06:56  <piscisaureus_>isaacs: merged it. no new failures.
19:07:03  <isaacs>awesome
19:07:10  <piscisaureus_>it was easy :-)
19:07:17  <piscisaureus_>but you never know
19:07:33  <isaacs>next week: let's get to 0 lint warnings and 0 test failures. moving tests out of the way is fine.
19:07:55  * travis-cijoined
19:07:55  <travis-ci>[travis-ci] joyent/libuv#95 (master - 1d942e2 : Bert Belder): The build is still failing.
19:07:55  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/4e1f2b1...1d942e2
19:07:55  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/672652
19:07:55  * travis-cipart
19:08:29  <isaacs>it's blocking contributions, because people are afraid of breaking stuff.
19:08:43  * `3rdEdenjoined
19:09:44  * mikealquit (Quit: Leaving.)
19:09:54  <piscisaureus_>they should be :-)
19:11:12  <isaacs>right, but they should be able to tell easily which thing they broke
19:11:23  <isaacs>it should be a china shop, not a rubble pile
19:11:37  <piscisaureus_>isaacs: :-)
19:12:30  * dshaw_joined
19:13:09  * brsonjoined
19:15:37  <igorzi>isaacs: what do you think about back-porting https://github.com/joyent/node/commit/c1a63a9 into v0.6 ?
19:16:39  <isaacs>igorzi: i saw the discussion on that.
19:16:46  <isaacs>igorzi: how was this done in v0.4?
19:16:56  <piscisaureus_>starttls?
19:16:59  <isaacs>i str that we had some kind of "socket" arg
19:17:04  * mikealjoined
19:18:19  <igorzi>isaacs: hmm, i don't know.. was tunneling possible in v0.4 ?
19:18:32  <isaacs>maybe i'm misremembering.
19:19:51  * mikealquit (Client Quit)
19:20:52  <isaacs>igorzi: who needs this? how badly?
19:21:57  <bnoordhuis>we should split node_crypto.cc in crypto and tls files sometime...
19:21:57  <igorzi>isaacs: can you skype for a few mins?
19:22:10  <isaacs>sure
19:25:48  * bnoordhuispart ("Leaving")
19:25:53  * bnoordhuisjoined
19:31:14  * philipsquit (Changing host)
19:31:14  * philipsjoined
19:32:57  * mjr_quit (Quit: mjr_)
19:32:59  * dshaw_quit (Ping timeout: 245 seconds)
19:36:37  * dshaw_joined
19:37:53  * AndreasMadsenchanged nick to c__
19:38:20  * c__changed nick to AndreasMadsen
19:38:25  * AndreasMadsenchanged nick to c__
19:38:35  * c__changed nick to AndreasMadsen
19:56:07  <CIA-110>node: Bert Belder reviewme * ref032cb / src/node.cc : Windows: support non-ansi command line arguments - http://git.io/_Oq9fA
19:56:14  <piscisaureus_>isaacs: igorzi: ^-- review?
19:56:56  <isaacs>piscisaureus_: that was fast
19:56:58  <isaacs>:)
19:57:10  <piscisaureus_>isaacs: that's only half of the fix
19:57:24  <piscisaureus_>isaacs: still need to do unicode environment
19:57:26  <isaacs>piscisaureus_: the other is for environments?
19:57:26  <isaacs>right
19:58:21  <igorzi>piscisaureus_: lgtm
19:58:30  <piscisaureus_>igorzi: that was fast :-)
19:58:30  * mikealjoined
19:58:45  * sh1mmerquit (Quit: sh1mmer)
19:59:05  <CIA-110>node: Bert Belder v0.6 * ref032cb / src/node.cc : Windows: support non-ansi command line arguments - http://git.io/_Oq9fA
19:59:28  <isaacs>piscisaureus_: lgtm
20:00:22  * sh1mmerjoined
20:04:06  * travis-cijoined
20:04:06  <travis-ci>[travis-ci] joyent/node#409 (reviewme - ef032cb : Bert Belder): The build was fixed.
20:04:06  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/commit/ef032cb
20:04:06  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/672928
20:04:06  * travis-cipart
20:07:31  * travis-cijoined
20:07:31  <travis-ci>[travis-ci] joyent/node#410 (v0.6 - ef032cb : Bert Belder): The build passed.
20:07:31  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/2e6ad62...ef032cb
20:07:31  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/672942
20:07:31  * travis-cipart
20:16:32  <CIA-110>node: isaacs master * r3a915b0 / (4 files in 3 dirs): Upgrade uv to 1d942e2a - http://git.io/CwXhSQ
20:17:04  <piscisaureus_>igorzi: should I use crt or win32 to manage environment variables?
20:17:29  <igorzi>piscisaureus_: i'd go with win32
20:18:18  <piscisaureus_>igorzi: you realize that if you change the enviroment using the win32 api, it gets out of sync with the crt api
20:18:32  <piscisaureus_>so changes you make may not be visible through getenv()
20:18:38  <piscisaureus_>igorzi: we think that's okay?
20:18:46  <piscisaureus_>we->you
20:19:02  <igorzi>piscisaureus_: do we use getenv() anywhere?
20:19:12  <piscisaureus_>igorzi: well - we will not longer do that
20:19:24  <piscisaureus_>igorzi: I guess it won't matter for compiled addons etc?
20:19:51  <piscisaureus_>we should actually put it in libuv
20:19:58  <igorzi>piscisaureus_: yep
20:24:20  <piscisaureus_>igorzi: bnoordhuis: isaacs: what are you guys working on atm (just so I stay in the loop).
20:24:44  <isaacs>piscisaureus_: i'm doing very boring documentation things mostly.
20:24:48  <piscisaureus_>I am going this unicode env first, then I'm going to comment on isaacs' domain proposals
20:24:56  <isaacs>piscisaureus_: and building 0.7.4 today
20:25:02  <piscisaureus_>kewl
20:25:29  <isaacs>considering yanking all the icons out of the dist tarballs and referring to them all by url, just to shut up this stupid legal discussion before it goes any further.
20:25:49  <piscisaureus_>all icons?
20:26:59  <igorzi>piscisaureus_: i'm trying to figure out how to do tunneling through HTTP using v0.6
20:27:18  <piscisaureus_>igorzi: what kind of tunneling?
20:27:22  <piscisaureus_>igorzi: SOCK proxy?
20:27:34  <creationix>isaacs, so they are concerned about the legal implications? I thought it was more the fact that the nodejs.org website doesn't need to be in the tarball
20:28:13  <isaacs>well.. actually, including the nodejs.org website in the tarball is kind of awesome, i think
20:28:33  <isaacs>but the images can be referred to via url.
20:29:11  <isaacs>hyperlinks are protected speech in many countries :)
20:29:27  <igorzi>piscisaureus_: HTTP proxy (through CONNECT)
20:29:39  <creationix>my only worry is bloating the tarball
20:29:58  <creationix>my nvm/src directory takes quite a bit of space on the harddrive
20:30:06  <creationix>but whatever
20:31:50  * travis-cijoined
20:31:50  <travis-ci>[travis-ci] joyent/node#411 (master - 3a915b0 : isaacs): The build is still failing.
20:31:50  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/5041bc1...3a915b0
20:31:50  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/673071
20:31:50  * travis-cipart
20:32:16  <isaacs>creationix: yeah, that, too. but these pngs are tiny, compared to the node binary.
20:32:22  <isaacs>(most of them, anyway)
20:32:31  <isaacs>restricting to a single color makes pngs really small :)
20:35:04  * AndreasMadsenquit (Remote host closed the connection)
20:39:30  * bentkusquit (Ping timeout: 244 seconds)
20:41:01  * toothrquit (Read error: Connection reset by peer)
20:41:14  * mikealquit (Quit: Leaving.)
20:42:02  <CIA-110>node: Fedor Indutny master * re7e4f2f / deps/v8/src/objects.cc :
20:42:02  <CIA-110>node: Fix linker error on some gcc version on osx
20:42:02  <CIA-110>node: * See: http://codereview.chromium.org/9382033/ - http://git.io/Hb3ikA
20:44:02  * toothrjoined
20:48:57  * piscisaureus_quit (Ping timeout: 276 seconds)
20:55:23  * mikealjoined
20:57:36  * travis-cijoined
20:57:36  <travis-ci>[travis-ci] joyent/node#412 (master - e7e4f2f : Fedor Indutny): The build is still failing.
20:57:36  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/3a915b0...e7e4f2f
20:57:36  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/673213
20:57:36  * travis-cipart
20:58:35  * isaacsquit (Remote host closed the connection)
21:10:51  * AndreasMadsenjoined
21:21:45  * indutnychanged nick to indutny_sleeping
21:25:44  * bbbbquit (Ping timeout: 252 seconds)
21:29:52  * sh1mmerquit (Quit: sh1mmer)
21:36:07  * piscisaureus_joined
21:36:54  * isaacs_mobilejoined
21:41:58  * indexzeroquit (Quit: indexzero)
21:42:54  * `3rdEdenquit (Quit: Linkinus - http://linkinus.com)
21:44:18  <CIA-110>libuv: Ben Noordhuis v0.6 * rb8649fd / src/unix/tty.c :
21:44:18  <CIA-110>libuv: unix: don't flush tty on switch to raw mode
21:44:18  <CIA-110>libuv: Drain, don't flush the tty when switching from cooked to raw mode. Prevents
21:44:18  <CIA-110>libuv: buffered keystrokes from getting lost. Switching back to cooked mode still
21:44:18  <CIA-110>libuv: flushes.
21:44:18  <CIA-110>libuv: Fixes joyent/node#2744. - http://git.io/GtZ24A
21:44:59  <piscisaureus_>bnoordhuis: ^-- that can't be fixed on windows, so I won't
21:45:06  <bnoordhuis>oh sure
21:45:13  <bnoordhuis>it's a minor issue anyway
21:45:43  <piscisaureus_>bnoordhuis: however if if we don't resume() stdin before switching to raw mode, all keystrokes that were buffered up so far will remain intact
21:45:50  * travis-cijoined
21:45:50  <travis-ci>[travis-ci] joyent/libuv#96 (v0.6 - b8649fd : Ben Noordhuis): The build is still failing.
21:45:50  <travis-ci>[travis-ci] Change view : https://github.com/joyent/libuv/compare/bc8b990...b8649fd
21:45:50  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/libuv/builds/673480
21:45:50  * travis-cipart
21:46:18  <piscisaureus_>keystrokes only get lost when we were actually reading in line-buffered mode
21:46:37  <bnoordhuis>piscisaureus_: 'we' in this context means 'windows'?
21:46:46  <piscisaureus_>we-ndows
21:48:21  * isaacs_mobilequit (Remote host closed the connection)
21:49:28  * dshaw_1joined
21:52:59  * dshaw_quit (Ping timeout: 245 seconds)
21:53:26  * isaacs_mobilejoined
21:55:11  * isaacs_m_joined
21:55:11  * isaacs_mobilequit (Read error: Connection reset by peer)
21:58:10  * isaacsjoined
22:04:26  * perezdjoined
22:04:41  * isaacs_m_quit (Remote host closed the connection)
22:30:19  * mikealquit (Quit: Leaving.)
22:35:59  * AvianFluquit (Quit: Leaving)
22:40:01  * AndreasMadsenquit (Remote host closed the connection)
22:41:41  <isaacs>piscisaureus_: how hard would it be to get your slurp thingie slurping #Node.js also?
22:41:48  <isaacs>piscisaureus_: http://nodejs.debuggable.com/ needs some redundancy :)
22:42:04  <piscisaureus_>isaacs: not hard - but right now it is a singleton
22:42:07  <isaacs>also, i think ircretary should probably do logging.
22:42:13  <isaacs>i had that once upon a time.
22:42:25  <isaacs>but it was... not good.
22:42:32  <isaacs>i'll put slurp in her brains
22:42:37  <piscisaureus_>unlike slurp which is great at logging
22:42:44  <piscisaureus_>but that's all it does
22:42:45  <piscisaureus_>https://github.com/piscisaureus/slurp
22:42:53  <piscisaureus_>I take patches
22:43:04  <piscisaureus_>I can also work on it tomorrow, should not be much work
22:43:33  * mikealjoined
22:43:36  <isaacs>meh. don't let it distract you.
22:44:30  <piscisaureus_>It's not difficult at all - it just needs to start multiple loggers
22:44:36  <piscisaureus_>or you would have to run it on another domain
22:47:27  <isaacs>i'd like to have ircretary slurp up all the channels she sits in
22:47:37  <isaacs>it'd be nice to see context when you get notes and mentions
22:47:47  <isaacs>they could link to the place in chat where it was actually said.
22:48:16  <piscisaureus_>yeah
22:48:27  <piscisaureus_>it would be nice, but it is also work :-)
22:48:32  <piscisaureus_>I considered this features
22:48:39  <piscisaureus_>also live socket.io logging, custom highlighting
22:48:50  <piscisaureus_>s/this/these
22:49:19  <piscisaureus_>but it is all relatively unimportant
22:49:38  <piscisaureus_>If I want anything for slurp it is the ability to regex search the entire log history
22:56:08  * mikealquit (Quit: Leaving.)
23:01:13  * mikealjoined
23:02:43  <isaacs>review, please? https://github.com/isaacs/node/commit/3c676032f24e3703aea4939b8dff975b24bb3ca8
23:02:46  <isaacs>mikeal: ^
23:04:38  * piscisaureus_quit (Ping timeout: 252 seconds)
23:08:24  <mikeal>awesome
23:08:32  <mikeal>i love commits that end list discussions :)
23:10:57  <mmalecki>isaacs: why urls?
23:11:06  <mmalecki>http://nodejs.org/images/icons-interior.png
23:11:24  <bnoordhuis>okay, disabling tls client renegotiation turned out to be harder than i thought
23:11:43  <mmalecki>isaacs: make them /images/..., so that browsers don't complain when using https
23:13:14  <isaacs>mmalecki: because it keeps them out of the dist tarballs, but makes them still work.
23:13:21  <isaacs>mmalecki: we can use // though, tha's a good point.
23:13:32  <mmalecki>isaacs: no, I mean, in css
23:13:43  <isaacs>right
23:13:46  <isaacs>that's what i mean.
23:13:52  <isaacs>url(//nodejs.org/images/blah.png)
23:13:59  * bnoordhuissigns off for today
23:14:00  <isaacs>same in src="..." and href="..."
23:14:03  <isaacs>bnoordhuis: g'nite
23:14:08  <bnoordhuis>you too
23:14:16  * bnoordhuisquit (Quit: Leaving)
23:14:57  <mmalecki>isaacs: or /images/..., that should work to, I think
23:15:00  <mmalecki>*too
23:15:09  <isaacs>mmalecki: not if you're looking at a html doc that isn't on the nodejs site.
23:15:26  <mmalecki>isaacs: ah
23:17:34  <isaacs>mmalecki: oh, actually, nevermind: https://nodejs.org/
23:17:38  <isaacs>does't work anyway :)
23:17:58  <mmalecki>oh? I thought it does
23:18:55  <isaacs>nope
23:19:34  <mmalecki>yeah, I mean, korsakov XD
23:20:11  <mmalecki>actually, it'd be cool if you guys made it work
23:20:18  <mmalecki>cause, like, downloads and stuff.
23:28:18  <CIA-110>node: isaacs master * ra7af84f / (90 files in 8 dirs):
23:28:18  <CIA-110>node: docs: Move images out of the dist tarball
23:28:18  <CIA-110>node: This puts all images in doc/images/ and references them via
23:28:18  <CIA-110>node: http://nodejs.org/images/.
23:28:18  <CIA-110>node: Any complaints about copyright usage etc. can thus be node/joyent's
23:28:18  <CIA-110>node: problem, rather than the problem of a downstream distribution channel. - http://git.io/onBhcQ
23:28:19  <CIA-110>node: isaacs master * rdf236e7 / doc/api_assets/style.css : css: Restrict li fix to #apicontent only. - http://git.io/IiQXvQ
23:28:19  <CIA-110>node: isaacs master * r9a0d907 / doc/template.html : doc: Wrap API doc content in div#apicontent - http://git.io/ZI2P0A
23:28:47  <CIA-110>node: isaacs v0.7.4-release * rd3dd9ed / (6 files in 5 dirs): (log message trimmed)
23:28:47  <CIA-110>node: 2012.02.13, Version 0.7.4 (unstable)
23:28:47  <CIA-110>node: * Upgrade V8 to 3.9.5
23:28:47  <CIA-110>node: * Upgrade npm to 1.1.1
23:28:47  <CIA-110>node: * build: Detect host_arch better (Karl Skomski)
23:28:47  <CIA-110>node: * debugger: export `debug_port` to `process` (Fedor Indutny)
23:28:47  <CIA-110>node: * api docs: CSS bug fixes (isaacs)
23:31:02  <CIA-110>node: isaacs v0.7.4-release * rf5ab8a7 / (6 files in 5 dirs): (log message trimmed)
23:31:03  <CIA-110>node: 2012.02.14, Version 0.7.4 (unstable)
23:31:03  <CIA-110>node: * Upgrade V8 to 3.9.5
23:31:03  <CIA-110>node: * Upgrade npm to 1.1.1
23:31:03  <CIA-110>node: * build: Detect host_arch better (Karl Skomski)
23:31:03  <CIA-110>node: * debugger: export `debug_port` to `process` (Fedor Indutny)
23:31:03  <CIA-110>node: * api docs: CSS bug fixes (isaacs)
23:41:16  * AvianFlujoined
23:42:57  * piscisaureus_joined
23:43:24  <piscisaureus_>Hmm when did google maps get so smart?
23:43:37  <piscisaureus_>It knows exactly where I am with gps and wifi both off
23:44:05  * travis-cijoined
23:44:05  <travis-ci>[travis-ci] joyent/node#413 (master - 9a0d907 : isaacs): The build is still failing.
23:44:05  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/e7e4f2f...9a0d907
23:44:05  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/673856
23:44:05  * travis-cipart
23:44:25  * travis-cijoined
23:44:25  <travis-ci>[travis-ci] joyent/node#414 (v0.7.4-release - d3dd9ed : isaacs): The build failed.
23:44:25  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/commit/d3dd9ed
23:44:25  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/673859
23:44:25  * travis-cipart
23:44:33  <isaacs>piscisaureus_: maybe your isp is selling you ou
23:44:34  <isaacs>t
23:44:51  <piscisaureus_>heh
23:45:18  <piscisaureus_>well I know in theory it would be possible to triangulate using the strengths of gps antennas
23:45:37  <piscisaureus_>But I didn't know that android could do that or that there even was an api for it
23:45:43  <isaacs>oh, you're on a cell phone, you mean?
23:45:52  <piscisaureus_>yeah
23:45:53  <isaacs>yes. cell tower latency is pretty good.
23:46:10  <isaacs>i mean, pretty measurable
23:46:25  <piscisaureus_>yeah
23:46:29  <isaacs>piscisaureus_: going to release 0.7.4 tomorrow morning.
23:46:40  * travis-cijoined
23:46:40  <travis-ci>[travis-ci] joyent/node#415 (v0.7.4-release - f5ab8a7 : isaacs): The build is still failing.
23:46:40  <travis-ci>[travis-ci] Change view : https://github.com/joyent/node/compare/d3dd9ed...f5ab8a7
23:46:40  <travis-ci>[travis-ci] Build details : http://travis-ci.org/joyent/node/builds/673862
23:46:40  * travis-cipart
23:46:42  <piscisaureus_>isaacs: alright, fine. anything you need in there?
23:46:57  <isaacs>piscisaureus_: nope. just review the changelog, see if i missed anything
23:47:05  <piscisaureus_>ok
23:47:15  <piscisaureus_>isaacs: I like this procedure
23:47:15  <isaacs>i've gotta hurry up and finish my homework before class.
23:47:23  <isaacs>yeah. it's a bit slower.
23:47:25  <piscisaureus_>isaacs: take some time to do a release.
23:47:35  <piscisaureus_>don't hurry and fuck it up anyway
23:47:38  <isaacs>yeah
23:47:48  <isaacs>the doc crap kind of took longer than i'd expected anyway
23:47:59  * mjr_joined
23:48:08  <piscisaureus_>isaacs: are you getting there with garen? It seems that you are not really in agreement?