[00:00] elliottcable: lol
[00:00] elliottcable: You’re in a JS channel, Tautologistics, we’re used to line termination o_O
[00:01] scudco has joined the channel
[00:01] Tautologistics: No. Terminators change based on context
[00:01] elliottcable: oh lol
[00:01] Tautologistics: in erlang it could be nothing, ".", or ";"
[00:12] Tautologistics has joined the channel
[00:30] alexiskander has joined the channel
[00:36] elliottcable: FFFF
[00:36] elliottcable: “KeyError: 'VENDOR'”
[00:36] dnolen has joined the channel
[00:40] elliottcable: K, well, I did my best. Maybe I can get ryah_away to take a look at it later.
[00:41] elliottcable: 3 hours is about my hard limit for staring blankly at C++ and Python build tools that I don’t understand.
[00:41] elliottcable: anybody know if ryah_away prefers GitHub messages or e-mail or what?
[00:42] jed has joined the channel
[00:45] isaacs: elliottcable: send to the list
[00:45] isaacs: what's going on?
[00:48] elliottcable: isaacs: After a few hours of poking around in the scons/waf/whatever tools, I got this far: http://github.com/elliottcable/node/commit/11035f6da339f922328225fdc22dd471d6d342e7
[00:48] elliottcable: isaacs: unfortunately, that blows up on `make` with a cryptic error.
[00:48] elliottcable: http://gist.github.com/254642
[00:50] elliottcable: afact the problem is that I duplicated the PREFIX setup here: http://github.com/elliottcable/node/blob/11035f6da339f922328225fdc22dd471d6d342e7/src/node_version.h.in#L12-13
[00:50] elliottcable: but build blows up on that file, so assumably the @VENDOR@ and @SITE@ are not getting replaced in-line. And I have no idea where in the clusterfuck-that-is-wafadmin the system is instructed to replace @PREFIX@ with the actual prefix as passed to --prefix
[00:50] elliottcable: >,<
[00:52] isaacs: yep, it's beyond me
[00:54] elliottcable: )-:
[00:54] elliottcable: god, I hate mailing lists.
[00:54] elliottcable: 
ACTION starts poking around, tyring to figure out how to get this information onto there, in such a way as he’ll actually see any helpful replies…

[00:55] isaacs: you should join the nodejs list
[00:55] isaacs: do it in digest mode, scan the headlines every few days.
[00:55] isaacs: it's a good practice.
[00:59] jcrosby has joined the channel
[01:03] aurynn: I just read everything
[01:06] elliottcable: inimino: I’m on the list
[01:06] elliottcable: aw shit, he left
[01:06] elliottcable: aurynn: too much for me. I’m on quite a lot of mailing lists, they all get filtered to the sidelines.
[01:06] alex-desktop has joined the channel
[01:06] elliottcable: 
ACTION absolutely hates mailing lists

[01:07] elliottcable: anyway, posted.
[01:07] elliottcable: If any of you have any ideas, please post: http://groups.google.com/group/nodejs/browse_thread/thread/cf5ae41c5add16ac
[01:09] jan____: deferreds!
[01:11] alexiskander has joined the channel
[01:11] aurynn: 
ACTION defers her opinion

[01:11] elliottcable: lol
[01:12] davidsklar has joined the channel
[01:12] elliottcable: 83 fricking nicks
[01:12] elliottcable: this place has grown o_o
[01:12] logicuce has joined the channel
[01:13] rictic has joined the channel
[01:30] jcrosby has joined the channel
[01:36] gwoo has joined the channel
[01:46] jed has joined the channel
[01:49] gwoo has joined the channel
[01:56] logicuce has left the channel
[02:13] rtomayko has joined the channel
[02:18] shfx has joined the channel
[02:19] Yuffster has joined the channel
[02:28] jazzychad has joined the channel
[02:37] jtaby has joined the channel
[02:38] dnolen_ has joined the channel
[02:41] jed has joined the channel
[02:43] binary42 has joined the channel
[02:47] RayMorgan has joined the channel
[02:56] eddanger has joined the channel
[03:29] Yuffster has joined the channel
[03:32] jed: tlrobinson: it's mostly just a stub, but i took a first rough cut at a JSGI wrapper for node.js.
[03:32] jed: tlrobinson: http://github.com/jed/http-jsgi/
[03:33] aurynn: This is awesome :)
[03:33] jed: i'd really like to see node.js standardize on the easier parts of the spec.
[03:33] jed: such as replacing req.uri with the JSGI equivalents.
[03:34] jed: i'm not too sure how async fits, so i left it out for now.
[03:34] aurynn: the best part of a new platform is, you can build all the libraries you like
[03:34] jed: yeah, the low-levelness is definitely motivating.
[03:35] jed: tlrobinson
[03:35] aurynn: It probably won't standardize on JSGI - there'll be a better metaphor soon
[03:35] aurynn: Someone ported WebMachine earlier today
[03:35] jed: tlrobinson's approach to promises is, as i understand it, to return a promise instead of a response, where the callback returns the actual response.
[03:36] jed: i don't know much about webmachine, will have to dig in.
[03:39] jed: aurynn: does webmachine support streaming?
[03:39] jed: aurynn: that's my biggest concern with JSGI.
[03:39] aurynn: I don't know if nodemachine does streaming
[03:42] sztanphet has joined the channel
[03:46] jtaby_ has joined the channel
[03:51] [i]chuck has left the channel
[03:54] JoePeck has joined the channel
[04:02] jed: aurynn: webmachine looks cool. watching the video now.
[04:02] aurynn: :)
[04:02] jed: aurynn: (but no continuous streaming, at least in the erlang version)
[04:03] aurynn: the node version may be different
[04:08] jcrosby has joined the channel
[04:11] eddanger has joined the channel
[04:22] dnolen has joined the channel
[04:27] hassox has joined the channel
[04:37] dnolen has joined the channel
[04:39] quirkey has joined the channel
[04:45] jtaby has joined the channel
[04:55] bentomas has joined the channel
[05:10] jed has joined the channel
[05:10] inimino: 
ACTION grumbles about kids these days... not knowing how to use mailing lists

[05:12] eddanger has joined the channel
[05:12] bentomas: inimino: what are "they" doing wrong?  Maybe they just haven't learned!
[05:12] inimino: apparently IBM's lawyers sent Crockford a letter to ask if they could get an exception to the "no evil" clause
[05:12] inimino: bentomas: yes, I think so
[05:13] jan____: inimino: yeah, he sent him one.
[05:13] inimino: bentomas: I think it's just an issue of not knowing how to use a decent email client (not that there really are any) and/or using primarily Webmail
[05:13] jan____: s/him/them/
[05:13] inimino: yeah
[05:13] aurynn: inimino, that's awesome
[05:14] aurynn: inimino, also, there are no decent email clients.
[05:14] inimino: I know :)
[05:14] inimino: but getting one that sucks less is a start
[05:15] bentomas: I'll never understand why everyone doesn't use GMail
[05:15] aurynn: Well, it's not suitable for corporate emails.
[05:15] scudco has joined the channel
[05:16] jan____: bentomas: gmail doesn't work for me
[05:17] bentomas: huh, well, I stand corrected!
[05:18] inimino: 
ACTION does not use Gmail

[05:18] bentomas: what do you all use?  
[05:19] aurynn: Thunderbird 3
[05:19] jan____: IMAP + Procmail + Mutt + Mail.app
[05:19] aurynn: it's lurvely
[05:20] inimino: mostly mutt and Thunderbird
[05:20] inimino: unfortunately mutt's IMAP support never really works well for me
[05:21] jan____: I use mutt on Maildirs that my IMAPd serves to mail.app
[05:22] emyller: Gmail here. (+ Google Apps)
[05:22] inimino: yeah I use mutt over SSH
[05:23] bentomas: huh, never used Mutt before, but I'll give it a try
[05:23] bentomas: what are its advantages?
[05:25] inimino: it's unequaled for dealing with large volumes of mail
[05:27] bentomas: inimino: was your complaint about "kids these days" targeted at anything in particular?  this is the first time I have participated in a mailing list and am curious about what are the faux pas
[05:28] inimino: bentomas: no, I was just remarking on someone complaining about mailing lists in the backscroll
[05:29] bentomas: ahh!  thanks for the help.
[05:29] inimino: bentomas: though I do notice a fair amount of top-posting and people not trimming their replies on the mailing list
[05:30] bentomas: inimino:  yeah not trimming replies is a major pet peeve of mine
[05:31] inimino: likewise
[05:50] Micheil has joined the channel
[06:10] aurynn: So to determine the byte size of a string, I walk the length of the string using charCodeAt and sum it?
[06:10] aurynn: hm
[06:10] aurynn: no, that isn't right
[06:10] jed: there's a method off of process for that.
[06:11] aurynn: which one?
[06:11] Tautologistics has joined the channel
[06:11] jed: aurynn: process._byteLength
[06:11] inimino: _byteLength or something
[06:11] inimino: but you would need to walk it and do UTF-16-to-UTF-8 math otherwise
[06:12] inimino: (depending on how you got the string)
[06:12] aurynn: _byteLength was it
[06:12] aurynn: I assume that it handles the conversion for me? :)
[06:12] jed: aurynn: yep.
[06:13] jed: aurynn: you should be good.
[06:13] aurynn: Awesome. Why isn't that in the doc?
[06:13] scudco has joined the channel
[06:13] aurynn: jed, thank you :)
[06:13] jed: aurynn: (i've used this too: encodeURIComponent( str ).split( /%./).join("").length )
[06:13] inimino: well, assuming you have a UTF-16 string
[06:13] jed: aurynn: no problem. i assume the underscore means they don't want it public?
[06:13] inimino: normally in node what you want to do is avoid doing that kind of encoding/decoding in the first place
[06:14] inimino: i.e. if the string is coming from a file you could use "binary" encoding to read it in and then you can just use .length
[06:15] aurynn: yeah, _ generally means "private". But then why is it exported?
[06:15] inimino: it's useful sometimes
[06:15] jed: aurynn: i'd imagine someone probably asked for it to be exposed.
[06:16] jed: aurynn: also, maybe it's the kind of thing better done in c++ than in javascript.
[06:19] aurynn: Maybe
[06:27] micheil_ has joined the channel
[06:33] bentomas has left the channel
[06:34] mahemoff has joined the channel
[06:35] Micheil has joined the channel
[06:41] micheil__ has joined the channel
[06:41] micheil__ has left the channel
[06:42] micheil__ has joined the channel
[06:42] micheil__: hmm..
[07:04] aurynn: Woo
[07:04] aurynn: My update of postgres.js still works for simple queries.
[07:31] hober has joined the channel
[07:39] scudco has joined the channel
[07:57] voodootikigod has joined the channel
[08:03] jcrosby has joined the channel
[08:05] rictic has joined the channel
[08:34] cmlenz has joined the channel
[08:52] simoncpu_ has joined the channel
[09:33] sveimac has joined the channel
[09:35] teemow has joined the channel
[10:31] tlrobinson: jed: cool i'll check it out
[11:30] jtaby has joined the channel
[11:46] saikko has joined the channel
[12:02] pdelgallego has joined the channel
[12:47] soveran has joined the channel
[12:54] malkomalko has joined the channel
[12:55] michaelk^ has joined the channel
[13:01] soveran has joined the channel
[13:26] sveimac has joined the channel
[13:27] binary42 has joined the channel
[13:35] rgl has joined the channel
[13:35] rgl: hi
[13:35] rgl: is there a way to serve static files (html/css/js)?   eg. for quick demos.
[13:40] jasondavies has joined the channel
[13:46] ryah_away: rgl: there is an example in node_chat
[15:06] ryah: snwoing in lübeck!
[15:18] Tautologistics has joined the channel
[15:29] rgl: ryah, thx. I'll look.
[15:30] saikko has joined the channel
[16:04] qFox has joined the channel
[16:17] malkomalko has joined the channel
[16:17] voodootikigod has joined the channel
[16:17] jed has joined the channel
[16:17] brandon_beacher has joined the channel
[16:17] onar has joined the channel
[16:17] mies has joined the channel
[16:17] rgl has joined the channel
[16:23] damien has joined the channel
[16:31] RayMorgan has joined the channel
[16:34] alex-desktop has joined the channel
[16:38] bentomas has joined the channel
[16:38] Micheil has joined the channel
[16:49] scudco has joined the channel
[16:54] Tautologistics has joined the channel
[17:11] gwoo_ has joined the channel
[17:13] jasondavies has joined the channel
[17:15] rgl: ryah_away, humm the node_chat does not work with current master branche of nodejs. it says there is no process.fs.cat function
[17:38] mediacoder: rgl: its posix.cat() now, but subject to change :-)
[17:51] rgl: ah thx. it works now :)
[18:06] erichocean has joined the channel
[18:15] sudoer has joined the channel
[18:30] claudio has joined the channel
[18:31] claudio: Hey. Using the recente libxml.js, I've written a (working) XML-RPC client for node.js. Is anyone interested to see/audit it? (this is my first node.js module, so I don't knwo if I'm on the right track) :)
[18:33] Connorhd has joined the channel
[18:39] aurynn: claudio, awesome. :)
[18:40] claudio: thanks :)
[18:40] eddanger has joined the channel
[18:41] claudio: I suppose I should open a github account, now... pant.
[18:42] michaelk^ has joined the channel
[18:48] trevor has joined the channel
[18:58] cloudhead has joined the channel
[18:58] Tautologistics has joined the channel
[19:09] logicuce has joined the channel
[19:10] logicuce: cxx is not a valid task generator -> ['command-output', 'cmd_output', 'cc', 'cmd', 'subst', 'copy']
[19:10] logicuce: I get above error while trying to build node.js in Ubuntu 9.04
[19:10] logicuce: just downloaded from Git
[19:10] logicuce: :(
[19:12] mediacoder: logicuce: install g++
[19:13] dnolen has joined the channel
[19:13] logicuce: gcc won't do?
[19:13] rgl has joined the channel
[19:13] logicuce: ok looking into it
[19:13] mediacoder: should work
[19:14] logicuce: configure works correctlt
[19:14] logicuce: I get error while 'make' is run
[19:14] logicuce: configure should have raised issues if it was compiler related.. what say
[19:14] logicuce: ?
[19:15] logicuce: gcc is already installed
[19:15] mediacoder: i dunno :-) i just rememebr reading this on the mailinglist
[19:16] logicuce: hmm
[19:16] logicuce: found this
[19:16] logicuce: http://www.travisswicegood.com/index.php/2009/07/11/compiling-node-js-on-ubuntu-9-04
[19:16] aurynn: Hm.
[19:16] aurynn: I suspect I'm generating bad queries.
[19:16] mediacoder: logicuce: http://groups.google.com/group/nodejs/browse_thread/thread/28b1fcdaae1b5ade
[19:16] ashb: build-essentials installed?
[19:18] logicuce: ashb: dunno
[19:18] logicuce: :(
[19:18] logicuce: lemme see
[19:18] logicuce: the link mediacoder gave talks about same problem though
[19:19] mediacoder: yea, and installing g++ seem to have solved it for tis guy..thats why i suggested it, too :-P
[19:20] logicuce: Thanks.. I am not that used to working in unix.. so taking my time to set up things
[19:20] logicuce: :)
[19:21] mediacoder: good luck :-)
[19:22] michaelk^ has joined the channel
[19:22] logicuce: So it seems build-essentials come up with g+
[19:22] logicuce: g++*
[19:22] logicuce: that should work... downloading
[19:22] logicuce: :) thanks guys
[19:27] jcrosby has joined the channel
[19:29] rgl has joined the channel
[19:30] isaacs has joined the channel
[19:42] logicuce: it worked and hello world also worked :)
[19:42] logicuce: Thanks again guys
[19:42] mediacoder: hehe, great :-)
[19:52] soveran has joined the channel
[20:07] aurynn: I'm trying to debug my network talking. Can I get what tcp is going to send out, in hex?
[20:09] inimino: aurynn: wireshark
[20:10] logicuce: if on windows then use MS Netmon
[20:10] aurynn: on OSX.
[20:18] aurynn: Oh, does Node have an objectDestroyed event?
[20:23] teemow has joined the channel
[20:27] sifi has joined the channel
[20:28] sifi: anyone know of a good commandline parser in js?
[20:28] inimino: aurynn: what is that?
[20:29] aurynn: inimino, it would allow me to implement cleanup
[20:29] aurynn: Basically analogous to sub DESTROY in perl, or __destroy__ in Python
[20:30] inimino: ah, no there is nothing like that in JavaScript
[20:30] aurynn: Right, which is why I wondered if node had an objectDestroyed event I could listen for
[20:30] aurynn: myObj.addListener("objectDestroyed", someCleanupFunction);
[20:30] inimino: I'm not sure how or why you'd implement that in a language that doesn't have it...
[20:31] aurynn: In this case, so I can clean up remote resources
[20:31] inimino: adding a hook to the garbage collector and checking callback functions on every collected object wouldn't be terribly efficient
[20:32] inimino: can you just do the cleanup you need directly?
[20:33] aurynn: Probably
[20:34] aurynn: 
ACTION goes back to hacking on postgres-js

[20:36] Tautologistics: aurynn: may be too late but two good choices for peeking at network traffic are ngrep and wireshark
[20:36] michaelk^ has joined the channel
[20:36] Tautologistics: Ah, wireshark's been mentioned
[20:37] isaacs has joined the channel
[20:38] mahemoff has joined the channel
[20:45] aurynn: hm
[20:46] rbranson_: unfortunately it's unlikely your long-lived objects will EVER be garbage collected in V8, unless you've got some serious churn
[20:47] aurynn: so the V8 process will just grow?
[20:47] rbranson_: it's very hesistant to actually GC
[20:47] rbranson_: and most of the objects you're using to handle a single request will get allocated in the 'young' generation
[20:48] aurynn: 
ACTION rewrites postgres-js to be hesitant with its network traffic

[20:48] rbranson_: what most people want is an event when the refcount hits 0
[20:48] rbranson_: and V8 doesn't really count references
[20:55] aurynn: Smarter GC than refcounting?
[20:56] rbranson_: yeah, it traverses the object graph at GC time
[20:56] alex-desktop: aurynn: Ref counting doesn't work if there can be circular references.
[20:56] rtomayko has joined the channel
[20:56] rbranson_: it can, but it gets tricky and expensive
[20:56] alex-desktop: yeah
[21:08] aurynn: It's so much fun to take someone's code, and rearchitect it completely.
[21:17] ryah: people really want autoreloading code...
[21:17] ryah: someone should build that into their framework
[21:21] mediacoder: hm, there is one project on github
[21:24] qFox: i dont suppose node can be "easily" installed and run adjecent to an existing lamp setup?
[21:24] qFox: on centos, no less :/
[21:24] qFox: or windows, but i understand that's going to be impossible for now
[21:25] qFox: like, have it run on a seperate port, not colliding with apache..
[21:25] qFox: but i suppose that'll cost too much configuration
[21:26] rbranson_: i guess it depends on what you consider too much :)
[21:27] qFox: i guess that answers my question :)
[21:27] qFox: yeah i'm a bit alergic too cli-ing and having to setup everything, yknow. i just want to code
[21:27] qFox: allergic
[21:28] rbranson_: node itself is pretty easy to install
[21:28] qFox: right now i'm not even sure whether my server has a python interpreter at all
[21:28] rbranson_: it's a ./configure; make; sudo make install
[21:29] qFox: rbranson_> sure, but will it screw up an existing apache config hosting on port 80?
[21:29] rbranson_: no
[21:29] qFox: then how does it serve
[21:29] rbranson_: most of the time people setup a proxy
[21:29] qFox: 
ACTION chuckles

[21:29] rbranson_: so that, say, /node goes to port 8888, where your node is runnig
[21:30] aurynn: Yay, proxay.
[21:30] rbranson_: but you can just connect to it at the port it's listening
[21:30] qFox: when i install it, what exactly happens? i mean, will it start running a daemon or anything?
[21:31] aurynn: think of node as an interpreter.
[21:31] aurynn: like python, or perl
[21:31] qFox: sure, but i also think of it as a webserver
[21:31] qFox: which is my concern..
[21:31] aurynn: It's not
[21:31] aurynn: It's an interpreter/runtime
[21:31] aurynn: upon which you can run webservers
[21:32] rbranson_: yeah, your node script itself says "I am a web server, listen on this port"
[21:32] rbranson_: not anything else
[21:32] qFox: node acts as an interface to serve content?
[21:32] rbranson_: nope, your code does
[21:32] qFox: (forgive my wording :)
[21:32] qFox: yeah, but through node
[21:32] rbranson_: node.js just provides the ability to read files, listen on ports, etc
[21:32] aurynn: Node is a programming environment
[21:32] qFox: i mean, node has to have the internals to be able to serve the content externally
[21:33] aurynn: Node is just a programming environment.
[21:33] qFox: hm
[21:33] aurynn: You can write code that talks TCP and serves HTTP content
[21:33] qFox: okay
[21:33] qFox: that makes sense (to me)
[21:33] aurynn: Or you can write code that walks your filesystem and does interesting things, but never touches the TCP interface
[21:34] qFox: anyone know whether it'll give problems on centOS?
[21:34] qFox: and how i can test for a python interpreter :p
[21:39] qFox: hmk
[21:47] mediacoder: qFox: i have it running on centOS
[21:48] mediacoder: no problems, might need newer python maybe
[21:52] aurynn: Implementing the postgres protocol is taking some major brainpower.
[21:53] qFox: mediacoder> how do i check whether the interpreter is of the correct version?
[21:53] qFox: and how do you update it... i'm already pretty sure it wont be up to do
[21:53] qFox: date.
[21:53] mediacoder: try installing node/running waf :-)
[21:54] qFox: i'm rather afraid installing it will b0rk something :/
[21:54] qFox: something i wont be able to fix myself
[21:54] mediacoder: node wont..probably if you mess too much with python
[21:55] qFox: ok but it will warn beforehand if python is too old, right?
[21:56] mediacoder: yea, it wont run the waf stuff and wont install anything :-)
[21:56] qFox: hmk
[21:56] qFox: i'll give it a go then
[21:57] RayMorgan has joined the channel
[22:22] kriskowal has joined the channel
[22:29] sudoer has joined the channel
[22:53] gwoo has joined the channel
[23:31] Tautologistics has joined the channel
[23:40] aurynn: should I do
[23:40] aurynn: var stmt = db.prepare("query")
[23:40] aurynn: or db.prepare("query").addCallback(function working with the statement);
[23:40] aurynn: ?
[23:44] aurynn: I'm leaning towards var stmt = 
[23:45] eddanger has joined the channel
[23:49] saikko has joined the channel
[23:49] inimino: aurynn: it requires a TCP roundtrip to the server to prepare the statement?
[23:50] aurynn: It requires 3 to do the prepare properly
[23:50] inimino: seems that should be asynchronous then, no?
[23:50] aurynn: yes
[23:50] aurynn: I'm trying to figure out how
[23:51] inimino: oh
[23:51] aurynn: I can't find a good place to hang the asynchronous bits off of
[23:51] dnolen has joined the channel
[23:51] inimino: well var stmt = db.prepare() should be fine
[23:52] inimino: then when it is first used, it will just have to wait for the prepare to come back
[23:52] inimino: is the only thing you can do with a prepared statement bind/execute it?
[23:52] rictic has joined the channel
[23:52] aurynn: Pretty much, yes.
[23:53] aurynn: Though theoretically you can hand it around, if you wanted
[23:53] aurynn: var stmt = db.prepare("query").addCallback(function (stmt){ return stmt; });
[23:53] aurynn: ;)
[23:54] rbranson_: stmt won't be the actual statement
[23:54] rbranson_: since it's async
[23:54] rbranson_: I don't think addCallback returns anything
[23:54] inimino: so there will be something like stmt.execute( ...bindings..., callback)
[23:54] inimino: it returns the promise itself
[23:55] rbranson_: well, the prepare should I assume as well?
[23:55] inimino: anyway, I don't think .prepare() needs to expose any callback
[23:55] inimino: the only thing you can do with a prepared statement is use it, right?
[23:55] inimino: so just expose the API for using it
[23:55] inimino: the use of it is going to be async
[23:57] inimino: so when it's used, it the prepare call to the db engine hasn't come back yet, you just wait until it has to send the next command
[23:57] inimino: but you don't need to expose the 'prepare call came back from the db' to the API user
[23:57] rbranson_: I see what you're saying
[23:58] rbranson_: any subsequent calls can wait until it's actually prepared
[23:58] aurynn: subsequent calls get buffered until it's prepared
[23:58] inimino: yes
[23:58] rbranson_: brilliant :D