00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:02:05  * felixgequit (Quit: felixge)
00:19:09  * AndreasMadsenquit (Remote host closed the connection)
00:30:12  * qmxchanged nick to qmx|away
00:32:02  <MI6>joyent/node: isaacs created branch v0.9.5-release - http://git.io/Mw36Ug
00:36:11  * kohaijoined
00:43:10  * c4milojoined
00:53:52  * c4miloquit (Remote host closed the connection)
00:55:17  <MI6>joyent/node: isaacs v0.9.5-release * 01994e8 : 2012.12.30, Version 0.9.5 (Unstable) * assert: improve support for new e (+1 more commits) - http://git.io/enbdYw
00:55:42  * c4milojoined
01:00:09  * c4miloquit (Ping timeout: 260 seconds)
01:18:11  * xaqjoined
01:25:14  <MI6>joyent/node: isaacs created tag v0.9.5 - http://git.io/tawzEw
01:25:53  <MI6>joyent/node: isaacs master * 30bd774 : Now working on 0.9.6 (+3 more commits) - http://git.io/p2B-KQ
01:27:26  <MI6>joyent/node: isaacs v0.8 * 44ceb5d : blog: Post for v0.9.5 - http://git.io/jGT6RA
01:41:08  * pooyajoined
01:48:30  * pooyaquit (Quit: pooya)
01:54:14  * xaqquit (Remote host closed the connection)
01:55:47  * xaqjoined
02:19:55  * xaqquit (Remote host closed the connection)
02:43:12  * c4milojoined
02:47:29  * c4miloquit (Ping timeout: 260 seconds)
03:40:50  * ryahjoined
03:42:38  <ryah>v0.9.5 very nice
03:43:02  <ryah>node-v0.9.5-linux-x64% ldd bin/node linux-vdso.so.1 => (0x00007fff835ff000) libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00007f33b4fee000) librt.so.1 => /lib/x86_64-linux-gnu/librt.so.1 (0x00007f33b4de5000) libstdc++.so.6 => /usr/lib/x86_64-linux-gnu/libstdc++.so.6 (0x00007f33b4ae5000) libm.so.6 => /lib/x86_64-linux-gnu/libm.so.6 (0x00007f33b47e9000)
03:43:07  <ryah> libgcc_s.so1 => /lib/x86_64-linux-gnu/libgcc_is.so.1 (0x00007f33b45d2000) libpthread.so.0 => /lib/x86_64-linux-gnu/libpthread.so.0 (0x00007f33b43b5000) libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007f33b3ff6000) /lib64/ld-linux-x86-64.so.2 (0x00007f33b520c000)
03:43:13  <ryah>pretty reasonable
03:43:20  <ryah>why is libstdc++.so.6 => /usr/lib/x86_64-linux-gnu/libstdc++.so.6 (0x00007f33b4ae5000)
03:43:24  <ryah>needed?
03:43:33  <ryah>does v8 depend on it?
03:49:22  <ryah>it's probably just linked in by default. i dont think it's used in node...
03:49:40  <ryah>oh well. bnoordhuis --^ maybe you know
03:49:50  <mmalecki>ryah: I know that libstdc++ is used by V8 on android for sure, not idea about x64
03:49:50  <ryah>not that it matters
03:49:58  <ryah>mmalecki: yeah?
03:51:41  * bnoordhuisquit (Ping timeout: 245 seconds)
03:55:44  <mmalecki>ryah: yeah, V8 is compiled with c++ which on my system links libstc++ in by default
04:04:38  * ryahquit (Quit: leaving)
04:29:59  * c4milojoined
04:34:06  * c4miloquit (Ping timeout: 244 seconds)
05:09:48  * AvianFlu_joined
05:11:07  * stagasquit (Ping timeout: 260 seconds)
05:12:28  * AvianFluquit (Ping timeout: 265 seconds)
05:36:58  * toothrjoined
05:37:02  * c4milojoined
05:51:05  * c4miloquit (Remote host closed the connection)
06:01:10  * AvianFlu_quit (Remote host closed the connection)
06:15:59  * c4milojoined
06:17:44  * ryahjoined
06:17:54  * ryahquit (Client Quit)
06:20:24  * c4miloquit (Ping timeout: 260 seconds)
07:06:16  * pooyajoined
07:21:01  * mikealquit (Quit: Leaving.)
07:25:10  * felixgejoined
07:28:43  * mikealjoined
07:29:43  * felixgequit (Ping timeout: 260 seconds)
07:32:23  * joshthecoderquit (Quit: Leaving...)
07:42:25  * AvianFlujoined
07:52:57  <txdv_>CAPSLOCK ATTACK
07:52:57  <LOUDBOT>HOW DID I KNOW BEFORE I OPENED THIS WINDOW THAT "ASS" WAS WHAT YOU SAID TO ME?
08:53:47  * c4milojoined
09:00:44  * AvianFluquit (Remote host closed the connection)
09:08:22  * c4miloquit (Remote host closed the connection)
09:13:43  * AvianFlujoined
09:14:47  * benoitcquit (Excess Flood)
09:16:47  * benoitcjoined
09:19:29  * hzjoined
09:31:44  * AvianFluquit (Remote host closed the connection)
09:51:33  * c4milojoined
09:56:00  * c4miloquit (Ping timeout: 264 seconds)
09:58:40  * rendarjoined
10:19:02  * benoitcquit (Excess Flood)
10:23:48  * benoitcjoined
10:58:01  * stagasjoined
11:07:11  * pooyaquit (Quit: pooya)
11:38:11  * c4milojoined
11:42:59  * c4miloquit (Ping timeout: 260 seconds)
11:50:50  * benoitcquit (Excess Flood)
11:51:50  * benoitcjoined
12:02:27  * felixgejoined
12:02:27  * felixgequit (Changing host)
12:02:27  * felixgejoined
12:22:29  * felixgequit (Quit: http://www.debuggable.com/)
12:27:18  * TheJHjoined
13:05:06  * bnoordhuisjoined
13:24:02  * c4milojoined
13:26:43  * benoitcquit (Excess Flood)
13:27:51  * benoitcjoined
13:28:20  * c4miloquit (Ping timeout: 244 seconds)
13:39:41  * bnoordhuisquit (Ping timeout: 255 seconds)
13:49:18  * Raltjoined
14:02:57  * loladirojoined
14:12:05  * indexzerojoined
14:22:02  * loladiroquit (Read error: Connection reset by peer)
14:22:07  * loladiro_joined
14:41:56  * Raltquit (Remote host closed the connection)
14:43:47  * stagasquit (Quit: ChatZilla 0.9.89-rdmsoft [XULRunner 1.9.0.17/2009122204])
15:14:57  * kuebkjoined
15:17:32  <kuebk>hi
15:17:52  <kuebk>is this guide up to date
15:17:53  <kuebk>http://nodejs.org/api/addons.html#addons_hello_world
15:17:53  <kuebk>?
15:21:51  <txdv_>IS IT?
15:23:20  <kuebk>;>
15:23:58  * indexzeroquit (Quit: indexzero)
15:32:57  * janjongboomjoined
15:33:08  <janjongboom>morning
15:33:19  <janjongboom>so I have a question regarding child_process spawn on node 0.6.21-pre
15:33:38  <janjongboom>why don't I have a stderr when spawning like: spawn("ssh", args, { customFds: [-1, -1, 2] })
15:33:53  <janjongboom>typeof child.stderr is undefined
15:34:26  <janjongboom>docs say "If the child stdio streams are shared with the parent, then this will not be set." for both stdout, stderr, stdin; but I do have stdout and stdin, so that doesn't seem to be the case
15:34:56  * loladirojoined
15:36:15  <janjongboom>w/o the customFds I do get stderr but I don't get why
15:36:19  * loladiro_quit (Ping timeout: 260 seconds)
15:52:32  * kuebkquit (Ping timeout: 255 seconds)
15:55:42  * stagasjoined
16:00:54  * kuebkjoined
16:32:17  * benoitcquit (Excess Flood)
16:39:14  * toothrquit (Ping timeout: 255 seconds)
16:39:22  * benoitcjoined
16:42:56  * toothrjoined
16:45:45  * janjongboomquit (Quit: Page closed)
16:48:27  * kuebkquit
16:50:29  * AvianFlujoined
16:58:31  * Raltjoined
17:06:06  * loladiroquit (Quit: loladiro)
17:06:16  * AndreasMadsenjoined
17:21:38  * c4milojoined
18:04:56  <indutny>hohohoho
18:07:20  <txdv_>santa with some commits?
18:07:53  <indutny>nononono
18:07:55  <indutny>no commits
18:08:11  <indutny>still working on making dtrace ustack helper even more awesome
18:08:21  <indutny>right now it ignores some top-level functions
18:08:24  <indutny>probably because of GC
18:09:00  * joshthecoderjoined
18:09:21  <txdv_>Found my old laptops hdd
18:10:45  * pooyajoined
18:22:40  * felixgejoined
18:22:40  * felixgequit (Changing host)
18:22:41  * felixgejoined
18:24:03  * c4miloquit (Remote host closed the connection)
18:33:24  * Soarezjoined
18:34:27  * piscisaureus_joined
18:35:20  <txdv_>what happened to faio?
18:35:26  <indutny>it's landed
18:35:31  <indutny>parts of it
18:35:32  <indutny>piscisaureus_: hi
18:35:34  <txdv_>where?
18:35:36  <txdv_>like in libuv?
18:35:37  <indutny>txdv_: libuv
18:35:52  <piscisaureus_>it didn't really land in libuv
18:36:00  <piscisaureus_>just something very similar to it
18:36:03  <piscisaureus_>indutny: hi
18:36:48  <indutny>well
18:36:51  <indutny>not exact code
18:36:59  <indutny>but yeah
18:37:16  <txdv_>will it be deleted now?
18:37:21  <indutny>dunno
18:37:22  <indutny>ask ben
18:40:32  <txdv_>servicestack, reddit and github are working on common markdown parser
18:45:09  <indutny>and?
18:45:23  <txdv_>awesome news if you love markdownj
18:46:08  * c4milojoined
18:47:29  <piscisaureus_>txdv_: I don't think faio will be deleted, or parts of libuv
18:47:43  <txdv_>i mean the repo called faio
18:50:27  * c4miloquit (Ping timeout: 252 seconds)
18:51:22  <txdv_>nevermind
18:51:26  <txdv_>it can rot on github
18:51:26  <piscisaureus_>txdv_: ben owns that, but I don't see why he would delete it
19:01:14  <txdv_>so it is basically a seperate project
19:01:19  * toothrquit (Ping timeout: 260 seconds)
19:03:56  * toothrjoined
19:04:27  * c4milojoined
19:04:46  <rendar>hmm, what is faio?
19:04:55  <rendar>txdv_: ^
19:05:17  <txdv_>basically what libuv does for unices
19:05:40  * Raltquit (Remote host closed the connection)
19:05:49  <rendar>txdv_: like libev ?
19:06:07  <txdv_>the sexy lean sister of that fat chick
19:07:50  <txdv_>does nikhilm come to this channel?
19:11:48  <rendar>txdv_: lol, i see, do you have some link? can't find nothing if googling 'faio'
19:13:54  <txdv_>saerch on github
19:13:57  <txdv_>it is bnoordhuis project
19:14:05  <txdv_>www.github.com/bnoordhuis/faio
19:16:38  * mikealquit (Quit: Leaving.)
19:17:26  <piscisaureus_>ircretary: funny, udp clustering load balances quite good
19:17:26  <ircretary>piscisaureus_: I'm not sure what to do with that command. Ask for help in PM.
19:17:36  <piscisaureus_>ircretary: tell bnoordhuis funny, udp clustering load balances quite good
19:17:37  <ircretary>piscisaureus_: I'll be sure to tell bnoordhuis
19:17:48  <tjfontaine>lets all switch to reliable udp transports
19:18:54  <txdv_>yeah thats because you can switch udp_t's between loops with ease
19:19:01  <txdv_>Close it one loop, reopen it on the other
19:19:15  <txdv_>can you do that with an active tcp connection?
19:19:22  <txdv_>damn that gave me some API idea
19:20:44  <txdv_>on one loop*
19:20:49  * c4miloquit (Remote host closed the connection)
19:22:59  * loladirojoined
19:23:06  <piscisaureus_>so nice :-)
19:23:21  <piscisaureus_>120k messages/sec with 4 workers in my ubuntu vm
19:23:38  <txdv_>one one core?
19:23:44  <txdv_>what is the use of workers if you have one vm core
19:23:44  <piscisaureus_>no 4 cores
19:24:13  <txdv_>so 30k per core?
19:24:18  <piscisaureus_>although 2 of them are saturated with sending messages
19:27:19  <txdv_>Could not import extension sphinx.ext.mathjax (exception: No module named mathjax)
19:27:19  <txdv_>Does anyone know how to add tha thing? (I am trying to compile the uvbook)
19:30:02  <txdv_>damn only from the 1.0 version, why nikhilm, whyyyy
19:30:43  <mmalecki>piscisaureus_: I hope it's not locally
20:02:28  * c4milojoined
20:03:14  * bnoordhuisjoined
20:17:07  * indexzerojoined
20:19:13  * sj26quit (*.net *.split)
20:19:30  * joclekjoined
20:19:52  * sj26joined
20:22:51  * toothrquit (Ping timeout: 252 seconds)
20:24:56  * toothrjoined
20:32:00  * c4milo_joined
20:36:44  * c4milo_quit (Ping timeout: 260 seconds)
20:42:52  * bentkusjoined
20:42:55  <bentkus>damn
20:43:01  <bentkus>that irc client ate too much ram
20:45:34  * `3rdEdenjoined
20:49:34  * loladiroquit (Ping timeout: 260 seconds)
21:04:01  <piscisaureus_>why does dgram emit bind errors async on 0.8 but sync on master .... ?
21:08:24  <piscisaureus_>bnoordhuis: hey
21:08:29  <bnoordhuis>piscisaureus_: ho
21:08:45  <bnoordhuis>i don't know. does it?
21:08:48  <piscisaureus_>bnoordhuis: what woul uv_udp_send2 do, I think you are not understanding the problem
21:09:09  <piscisaureus_>bnoordhuis: we need a function that sends an udp handle over a pipe
21:09:17  <piscisaureus_>so that would be uv_write3 :-)
21:09:20  <bnoordhuis>ah right
21:09:35  <piscisaureus_>bnoordhuis: or - uv_write should take an uv_handle_t* instead of uv_stream_t*
21:09:46  <piscisaureus_>bnoordhuis: but I thought I'd not change the API too much in this stage
21:09:52  <bnoordhuis>maybe we should abandon that
21:10:08  <piscisaureus_>bnoordhuis: what?
21:10:09  <bnoordhuis>and have something like uv_handle_send(uv_pipe_t*, uv_handle_t*)
21:10:22  <piscisaureus_>ah - that's a possibility
21:10:22  <bnoordhuis>abandon in the future, i mean
21:10:39  <piscisaureus_>bnoordhuis: but really, I think making uv_udp a stream would be the best solution for the :-)
21:10:45  <bnoordhuis>no, no, no
21:10:48  <piscisaureus_>bnoordhuis: we could also DRY uv_tcp_bind and uv_udp_bind
21:10:55  <bnoordhuis>have you been talking to ryan again?
21:10:59  <piscisaureus_>bnoordhuis: just have uv_ip_bind(uv_stream_t*)
21:11:01  <piscisaureus_>bnoordhuis: not about this :-)
21:11:24  <piscisaureus_>bnoordhuis: but really I would want to merge uv_tcp_t, uv_tty_t, uv_pipe_t etc into a single stream class
21:11:27  <piscisaureus_>which is opaque to the user
21:11:32  <piscisaureus_>it's also much nicer for you :-)
21:12:06  <piscisaureus_>bnoordhuis: because you don't have to do all this duplication across pipe and tcp - on unix you deal with them in the same way anyway
21:12:25  <bnoordhuis>yeah. still, i don't really mind the way it is now
21:12:43  <bnoordhuis>wouldn't merging everything into a single struct make it really big on windows?
21:12:55  <piscisaureus_>bnoordhuis: possibly. But there are ways.
21:13:05  <bnoordhuis>that don't involve malloc'ing memory?
21:13:28  <piscisaureus_>bnoordhuis: well, not on the hot paths. Big structs currently are tcp server structs (they malloc anyway) and tty handles (but they are rare)
21:13:38  * loladirojoined
21:13:39  <bnoordhuis>hm, okay
21:13:52  <bnoordhuis>i could see the case for that (i.e. merging)
21:14:04  <bnoordhuis>but that's something for the deep future
21:14:06  <piscisaureus_>it also makes the extension space much nicer
21:14:31  <piscisaureus_>e.g. if someone would want to add support for serial ports or wicked /dev devices we wouldn't have to add another uv_xyz_t
21:14:37  <piscisaureus_>and a lot of new api space
21:15:17  <piscisaureus_>the bsd socket abstraction is quite good in many ways, I mean it merges local pipes and datagram sockets etc
21:15:19  <bnoordhuis>true. each new widget would change the abi though
21:15:24  <piscisaureus_>why?
21:15:33  <piscisaureus_>the internal abi maybe
21:15:42  <piscisaureus_>and it could also add some functionality
21:15:43  <bnoordhuis>at least, if it increases sizeof(uv_stream_t)
21:15:52  <piscisaureus_>yes - that sucks
21:16:38  <piscisaureus_>we could also just fix the size for it
21:16:43  <piscisaureus_>like, 256 bytes
21:16:47  <piscisaureus_>or 128
21:16:49  <piscisaureus_>whatever
21:16:49  <bnoordhuis>because that's all you ever need
21:16:54  <piscisaureus_>well
21:17:00  <piscisaureus_>if you need more, go malloc with ur ass
21:17:14  <piscisaureus_>ok - I'm not sure that's so great
21:17:16  <bnoordhuis>but $DEITY kills a kitten every time you malloc
21:17:23  <piscisaureus_>I hate kittens
21:17:40  <bnoordhuis>ah. uv-win starts to make a lot more sense all of a sudden
21:17:46  <piscisaureus_>bnoordhuis: so in that light I'd like uv_udp_t to also be an uv_stream_t but one that can't read but it can receive
21:18:11  <bnoordhuis>the reason why i think that's an incredibly bad idea is
21:18:22  <bnoordhuis>that streams are 1-on-1 communication
21:18:35  <bnoordhuis>but datagram sockets are 1-to-many
21:18:39  <bnoordhuis>or many-to-1
21:18:56  <bnoordhuis>the streams analogy breaks down fast with datagram sockets
21:19:11  <piscisaureus_>but we could have uv_ip_bind(uv_stream_t* stream, struct sockaddr* addr, int flags);
21:19:11  <piscisaureus_>and uv_path_bind(uv_stream_t* stream, char* path, int flags);
21:19:11  <piscisaureus_>and uv_open(uv_stream_t* stream, int fd); <-- no need to care about the type
21:19:54  <piscisaureus_>bnoordhuis: sure. It also breaks with server sockets, because they don't receive data at all. Yet uv_tcp_t can be a server and nobody gives a shit.
21:20:02  * loladiroquit (Read error: No route to host)
21:20:55  <bnoordhuis>piscisaureus_: apples, oranges. i'm not impressed with your counterargument :)
21:21:03  <bnoordhuis>there's also things to consider like peer address
21:21:37  <piscisaureus_>bnoordhuis: how's that apples and oranges. The thing is that all these struct types are annoying, not that we pretend UDP == 1-to-1 reliable data stream
21:21:58  <piscisaureus_>bnoordhuis: so - yes, an udp socket has no peer address. Neither has a named pipe.
21:22:05  <piscisaureus_>(or an unnamed pipe for that matter)
21:22:13  <bentkus>wow
21:22:14  <bentkus>so much chat
21:23:03  <bentkus>don't make udp a stream
21:24:00  <piscisaureus_>bentkus: why???
21:24:45  <bentkus>cause they have ipaddr and data
21:25:04  <bnoordhuis>piscisaureus_: unix sockets have (or can have) a peer but okay, i digress :)
21:25:49  <piscisaureus_>bentkus: I am not saying we will remove uv_recv and uv_send
21:25:54  <piscisaureus_>that was *not* the plan
21:27:04  <bnoordhuis>maybe we need something in between uv_stream_t and uv_udp_t
21:27:31  <bnoordhuis>though i don't know what that would be
21:27:40  <bentkus>leave udp alone
21:28:33  <bentkus>it has done nothing wrong to you
21:28:54  <bentkus>in manos they created a stream abstraction of udp
21:28:57  <bentkus>it was an abomination
21:29:04  <bentkus>I begged me to kill it
21:29:16  <bentkus>it*
21:31:20  <bentkus>Damn I abstracted too much...
21:34:21  <piscisaureus_>god dammit I DO NOT WANT TO ABSTRACT UDP
21:34:31  <piscisaureus_>I just want them to use the same struct types
21:34:42  <bentkus>O
21:34:43  <piscisaureus_>and not duplicate APIs where the implemnetation is the same anyway
21:34:43  <bentkus>ok, do that
21:34:54  <mmalecki>piscisaureus_: +1
21:35:04  <piscisaureus_>like, as I said, bind, and uv_accept, and write2
21:35:21  <mmalecki>I could use it in the nearest future, having similar APIs would be great
21:35:28  <bentkus>yeah that would be sexy, I misunderstood you first
21:35:43  <bentkus>mmalecki: how could you use it in the nearest future?
21:35:47  <bnoordhuis>so uv_accept is kind of meaningless for udp sockets
21:35:59  <bnoordhuis>only when passing around handles
21:36:52  <mmalecki>bentkus: I'm writing an app which can need both TCP and UDP and I'd rather now have to unify the API myself
21:41:27  <bentkus>in C?
21:42:19  <piscisaureus_>bnoordhuis: yes, so
21:42:31  <piscisaureus_>bnoordhuis: uv_accept(udp_handle, client) <-- UV_ENOTSUP
21:42:50  <piscisaureus_>bnoordhuis: uv_accept(pipe_handle, udp_handle) <-- get udp handle from ipc pipe
21:43:10  * rendarquit
21:43:13  <piscisaureus_>or whatever error - EPROTONOOPT or EOPNOTSUP
21:43:15  <bnoordhuis>piscisaureus_: i kind of see the logic of that if you turn its prototype into uv_accept(uv_handle_t*, uv_handle_t*)
21:44:12  <piscisaureus_>bnoordhuis: true, but then people still need to do
21:44:46  <mmalecki>bentkus: yeah
21:44:55  <mmalecki>sometimes you need to drop down to plain C :)
21:45:38  <piscisaureus_>void on_read2(uv_stream_t* pipe, int type) {
21:45:39  <piscisaureus_> switch (type) {
21:45:39  <piscisaureus_> case UV_UDP: {
21:45:39  <piscisaureus_> uv_udp_t* h = malloc(sizeof *h)
21:45:39  <piscisaureus_> uv_udp_init(h)
21:45:39  <piscisaureus_> uv_accept(pipe, h)
21:45:39  <piscisaureus_> break;
21:45:40  <piscisaureus_> }
21:45:40  <piscisaureus_> case UV_TCP: {
21:45:41  <piscisaureus_> ... etc
21:45:53  <piscisaureus_>^-- bnoordhuis: that was also a sucky compromise
21:46:42  * wolfeidauquit (Remote host closed the connection)
21:47:02  <bnoordhuis>piscisaureus_: you wanted to stuff everything into one big struct, right?
21:47:15  <piscisaureus_>bnoordhuis: well, yes
21:47:34  <piscisaureus_>bnoordhuis: not *everything* but everything that one could conceivably send over a pipe, yes
21:47:46  <piscisaureus_>bnoordhuis: it doesn't make much sense for prepare, asymc, timer etc
21:47:56  <piscisaureus_>and those structs are really a lot smaller than typical streams
21:48:32  <bnoordhuis>hmm, so you want something like a uv_io_t struct?
21:49:24  <piscisaureus_>yes
21:49:38  <piscisaureus_>but not as a *base class* like uv_stream_t and uv_handle_t
21:49:58  <piscisaureus_>instead, a struct is guaranteed to be big enough to hold whatever it can be
21:51:00  <piscisaureus_>bnoordhuis: but the user shouldn't have to deal with uv_pipe_t vs uv_tcp_t. It is all the same, except in very specific circumstances (e.g. when you are binding)
21:51:19  <piscisaureus_>but not in situations where the only difference is the size of the struct (like when accepting)
21:51:34  <bentkus>mmalecki: what are you writing?
21:51:49  <bnoordhuis>piscisaureus_: let me sleep on that. it doesn't sound all bad
21:52:07  <piscisaureus_>bnoordhuis: good :-)
21:52:43  <piscisaureus_>bnoordhuis: we should at some point have a beer and cut the knot (is that a meaninful expression in this weird lingua franca?)
21:52:57  <piscisaureus_>bnoordhuis: I will even make a full api proposal for you :-)
21:53:05  <bnoordhuis>piscisaureus_: okay, if you're buying
21:53:13  <piscisaureus_>bnoordhuis: hmm
21:53:29  <bnoordhuis>btw, i challenge you the ancient ritual of koehandel
21:53:36  <piscisaureus_>bnoordhuis: we can have a new rule. The guy who is in his home town buys :-p
21:53:40  <piscisaureus_>hmm maybe not
21:53:51  <bnoordhuis>sounds good to me, actually
21:54:02  <piscisaureus_>are you planning on coming to amsterdam often?
21:54:02  * mikealjoined
21:54:08  <bnoordhuis>now i am, yes
21:54:44  <piscisaureus_>hahaha
21:54:45  <piscisaureus_>shit
21:54:56  <piscisaureus_>we can try it out when I come to gouda
21:55:05  <piscisaureus_>kijken hoe het bevalt
21:56:45  <bnoordhuis>good thing beer is a bit cheaper here than it is in 020 :)
21:59:03  * c4miloquit (Remote host closed the connection)
22:02:30  <piscisaureus_>haha
22:02:37  <piscisaureus_>you should go to paris :-p
22:02:45  <piscisaureus_>you'll never complain about 020 again
22:05:00  * joshthecoderquit (Quit: Leaving...)
22:05:53  <piscisaureus_>bnoordhuis: this patch doesn't support sending dgram.Socket objects with send()
22:05:58  <piscisaureus_>bnoordhuis: shall I just remove that code path?
22:06:17  <bnoordhuis>piscisaureus_: if it doesn't work then yes, remove it
22:07:24  <mmalecki>Paris is the most expensive city I've ever been to
22:09:28  <bnoordhuis>mmalecki: try london
22:09:44  <bnoordhuis>though inner city paris is expensive too, yes
22:10:14  <mmalecki>bnoordhuis: hmm, as I recall it, Paris was more expensive. maybe I'm failing at currency conversion, as usual :)
22:10:33  <mmalecki>dinner for 3 in Paris was over 100 euros, that was crazy
22:11:34  <bnoordhuis>mmalecki: say 35 euro per person? that's not too shocking, is it?
22:11:36  <mmalecki>Amsterdam on the other hand, was quite nice in that regard :)
22:12:05  <bnoordhuis>if you go outside the city centre, it's usually a lot cheaper and nicer
22:12:06  <mmalecki>more like around 50
22:12:22  <mmalecki>but I got to eat next to Napoleon's hat, so worth it
22:15:26  <mmalecki>(the story is: Napoleon used to drink his ass off in the certain place, then left his hat in pawn)
22:16:31  * perezdjoined
22:16:35  * wolfeidaujoined
22:20:08  <bentkus>binch drinking
22:22:17  <bentkus>uv_tty_get_winsize is not a blocking call?
22:24:56  * `3rdEdenquit (Remote host closed the connection)
22:25:31  <piscisaureus_>it is a blocking call (but it should be fast)
22:25:44  * felixgequit (Quit: felixge)
22:28:43  <piscisaureus_>bnoordhuis: if this sticks I'll add a tests and start backporting to 0.8.
22:28:49  <piscisaureus_>(for private use)
22:28:57  <bnoordhuis>okay
22:29:54  <piscisaureus_>oh btw - i noticed something odd
22:30:08  <piscisaureus_>bnoordhuis: do you know this - do UDP sockets bind to loopback by default?
22:30:19  <bnoordhuis>piscisaureus_: i know and they don't
22:30:30  * jmar777joined
22:30:30  <piscisaureus_>bnoordhuis: odd thing, with cluster they do
22:30:48  <bnoordhuis>that's odd, yes
22:30:52  <piscisaureus_>bnoordhuis: unless I explicitly bind to 0.0.0.0, they will be bound to 255.255.255.255
22:30:56  <piscisaureus_>I will dig some more
22:31:06  <bnoordhuis>255.255.255.255? wut?
22:31:57  <piscisaureus_>yes
22:33:35  <piscisaureus_>bnoordhuis: shows up like this in netstat:
22:33:36  <piscisaureus_>udp 0 0 255.255.255.255:9123 *:* 15222/node
22:33:45  <bentkus>\033 in C is octal?
22:34:25  <piscisaureus_>(port 9123 is the port I bound to, and 15222 is the master process
22:34:34  <bnoordhuis>piscisaureus_: udp 0 0 0.0.0.0:31019 0.0.0.0:* 31307/node <- that's the expected behavior
22:34:38  <bnoordhuis>bentkus: yes
22:34:46  <piscisaureus_>bnoordhuis: that's what happens when I don
22:34:48  <piscisaureus_>er
22:34:54  <piscisaureus_>*when I don't use cluster
22:34:58  <bnoordhuis>when you don... your wizard hat and robe?
22:36:44  <piscisaureus_>ah hmm
22:37:47  <piscisaureus_>just seems to be some oddity in dgram.js
22:39:58  <piscisaureus_>bnoordhuis: nvm - my bad
22:41:31  <bnoordhuis>$ out/Release/node -e 'require("dgram").createSocket("udp4").send("BAM")'
22:41:31  <bnoordhuis>node: ../../src/udp_wrap.cc:220: static v8::Handle<v8::Value> node::UDPWrap::DoSend(const v8::Arguments&, int): Assertion `Buffer::HasInstance(args[0])' failed.
22:41:45  <bnoordhuis>that also happens without your patch though
22:41:49  <piscisaureus_>ye
22:41:52  <piscisaureus_>s
22:41:57  <piscisaureus_>bnoordhuis: because send doesn't take strings
22:42:06  <piscisaureus_>bnoordhuis: it's something we should support I suppose
22:42:10  <piscisaureus_>bnoordhuis: also, https://github.com/piscisaureus/node/commit/cc3bfa250aeeb729c413b8bcc8aaffc90459139d
22:42:27  <bnoordhuis>yes, probably. but it definitely shouldn't crash :)
22:42:33  <piscisaureus_>bnoordhuis: the UDP API is kinda rough :-)
22:42:45  <piscisaureus_>bnoordhuis: ah - true
22:42:57  <bnoordhuis>i'll open an issue
22:43:29  <bentkus>are you going to redesign the node udp api?
22:43:41  <piscisaureus_>no
22:43:48  <piscisaureus_>just polishing it here and there, mebbe
22:47:20  <piscisaureus_>ok, ff wat voer naar binnen werken
22:47:43  * piscisaureus_brb
22:53:28  * hzquit
22:57:36  * jmar777quit (Remote host closed the connection)
23:03:13  <bnoordhuis>piscisaureus_: btw, `make jslint` complains
23:28:42  <bentkus>JSLINT
23:28:48  <bentkus>i dont like jslint
23:37:06  * AndreasM_joined
23:37:31  <piscisaureus_>bnoordhuis: updated
23:39:45  * AndreasMadsenquit (Ping timeout: 252 seconds)
23:46:21  * AndreasM_quit (Remote host closed the connection)
23:55:19  * toothrquit (Ping timeout: 260 seconds)
23:55:53  <bnoordhuis>https://github.com/bnoordhuis/node/blob/v0.8/CONTRIBUTING.md <- review?
23:57:10  <mmalecki>bnoordhuis: I don't like the caps in headers
23:57:22  <bnoordhuis>why not?
23:57:36  <bnoordhuis>WHY DO YOU HATE CAPS MACIEJ?
23:57:37  <LOUDBOT>OUT OF NOWHERE "IT'S NOT RAEP GUISE"
23:58:20  <bnoordhuis>writing in all caps makes me feel young again somehow
23:58:20  <piscisaureus_>bnoordhuis: Wrap all other lines at 80 columns.
23:58:23  <piscisaureus_>72?
23:58:31  <mmalecki>yeah, 72 is the rule
23:58:44  <bnoordhuis>piscisaureus_: oh. i (or rather, vim) always wraps at 80
23:58:48  <piscisaureus_>bnoordhuis: also, CLA
23:58:52  <bnoordhuis>yeah, it's there
23:58:55  <mmalecki>bnoordhuis: my vim wraps at 72
23:59:10  <piscisaureus_>bnoordhuis: too small.
23:59:20  <piscisaureus_>***SIGN THE FUCKING CLA***
23:59:20  <LOUDBOT>TOO EXPENSIVE TO BALANCE THE CHECKBOOK HUH DOUCHEBAG
23:59:23  <bnoordhuis>haha
23:59:32  <bnoordhuis>SIGN IT. OR ELSE.
23:59:32  <LOUDBOT>I CAUGHT THE AMATEUR PORN ACTRESS STROKING HER PUSSY AFTER THE SHOOT