[00:02] mscdex: johnster: what type of data is this and what do you plan on doing with the data?
[00:02] johnster: Its a HTML page that is encoded in iso8859-1
[00:02] johnster: I need to modify the content
[00:03] johnster: ^^^ it's an HTML page
[00:05] mscdex: pheuter: here's a quick example of reading a line at a time: http://gist.github.com/526123
[00:06] pheuter: mscdex: ah, i see... thnx
[00:07] mscdex: johnster: well, there is a couple of html parsers for node. about the only thing i've ever done in this dept. is running tidy on the html (converting it to xml in the process) and then using an xml library like libxmljs to scrape data
[00:08] johnster: Hmm. The problem isn't so much parsing the HTML as just getting it into a normal JavaScript string format
[00:08] johnster: but thanks : P maybe i'll find something searching through those
[00:10] pheuter: mscdex: ah, it works perfectly, how is it that it keeps on looping? So, once it reachers an end of line, why doesn't the buffer clear?
[00:11] robotarmy has joined the channel
[00:11] mscdex: pheuter: i don't understand what you mean
[00:12] pheuter: buffer = data.substr(data.indexOf("\n")+1); what is the purpose of that
[00:13] mscdex: pheuter: just in case there are characters after the newline
[00:13] mscdex: pheuter: node reads in chunks
[00:13] mscdex: pheuter: although probably for most cases it will read in the entire line at once because it fits in the size of a single chunk
[00:14] pheuter: technically, shouldnt u clear the buffer, so that the next line i input doesnt get mixed in with the previous one!
[00:14] mscdex: pheuter: imagine if you cat'ed a file to the script
[00:14] pheuter: sorry about the !, lol
[00:15] mscdex: pheuter: no, then you could lose part of the next line
[00:15] pheuter: mscdex: hm, i changed the buffer = to buffer = "", and it still worked
[00:16] mscdex: pheuter: i.e., the variable data == 'foo\nbar' and the next time the event gets emitted data == 'baz\n'
[00:16] mscdex: pheuter: in that case, you would discard 'bar', and the next line wouldn't be complete
[00:17] herbySk has joined the channel
[00:17] mscdex: pheuter: that is why i have it check for anything after the newline and have that as the beginning of the next line
[00:17] pheuter: mscdex: got it, thnx
[00:17] _announcer: Twitter: "Nitrode: A lightweight HTTP Server on top of Node.js http://thechangelog.com/post/958501415" -- Fabien Penso. http://twitter.com/fabienpenso/status/21271161375
[00:19] mscdex: pheuter: but yeah like i said, you probably wouldn't run into that scenario by just entering things in the command line, but if you do like: "cat script.foo | node calculator.js", then you could easily see the scenario happen depending on the length of script.foo
[00:20] pheuter: mscdex: exactly
[00:20] mscdex: er not command line, but stdin
[00:21] _announcer: Twitter: "@BioNuc didn't read any node.js thing... Relied on my understanding for evented code in ruby instead!" -- Ahmed Sobhi. http://twitter.com/humanzz/status/21271416568
[00:22] _announcer: Twitter: "Non-blocking operations and deferred execution with node.js - http://bit.ly/bk8Jei" -- dierken. http://twitter.com/dierken/status/21271425753
[00:22] _announcer: Twitter: "node-DHT: Distributed hash table library for nodejs http://bit.ly/avLJA1 # # DHT nodejs" [vi] -- Régis Gaidot. http://twitter.com/rgaidot/status/21271471586
[00:23] rauchg_ has joined the channel
[00:24] pheuter has joined the channel
[00:27] mbleigh has joined the channel
[00:32] scoates: (yes, I know this is technically a bad strategy, but this is for a very specific non-server application) is there a way to intentionally block? I can do while (checkCondition()) {};  but that eats 100% CPU. I'd prefer to while(checkCondition()) { block(10); /* 10ms */ } ... any way to do this?
[00:34] mbleigh: scoates: couldn't you just do an interval every 10ms that runs the check?
[00:35] mbleigh: or a timeout that reinitializes itself
[00:35] scoates: no, because I don't want the enclosing function to return until the check succeeds
[00:35] mscdex: scoates: mind sharing why you want to do this?
[00:36] scoates: it's a bit complicated, but basically, I want to emulate synchronous http requests in webshell
[00:36] mscdex: you can do that without doing that
[00:36] mscdex: without blocking i mean
[00:37] Tim_Smart: I made a async cli prompt tool recently actually.
[00:38] scoates: I don't want async (-:
[00:38] scoates: mscdex: can you elaborate?
[00:38] mscdex: scoates: sure, 1 moment
[00:38] tilgovi has joined the channel
[00:39] jesusabdullah: Oh, I see what you mean
[00:40] bpot has joined the channel
[00:41] pheuter has joined the channel
[00:41] steadicat has joined the channel
[00:42] mscdex: here's an example from earlier, it synchronously fetches www.google.com 100 times and waits until the last one is finished: http://gist.github.com/526167
[00:43] _announcer: Twitter: "node.js a JavaScript front-end event I / O library seems http://nodejs.org/" [ja] -- Yuuki Takano. http://twitter.com/ytakanoster/status/21272783654
[00:43] scoates: ah.. /-: I had something similar to that, but I also have a REPL in the mix, which does nasty things to the output stream when it displays its prompt (triggered on request end), which is part of what I'm trying to avoid.
[00:44] scoates: I mean, thanks mscdex, but I'm not sure this will solve my problem without a bunch of rearchitecting.
[00:45] mscdex: i don't understand how that code would affect a REPL?
[00:47] scoates: I have a doHttpRequest method that on response end, does a displayPrompt on the REPL. Once that happens, I can't just output normally because the REPL does things to the cursor. I know this isn't very clear from my explanation.
[00:48] mscdex: yeah, i'd have to see a small example that reproduces this
[00:48] cheriot has joined the channel
[00:48] scoates: yep.. and it would take a bunch of boiling down.. and then some groking.. /-: I appreciate the example, anyway.
[00:48] mscdex: heh
[00:49] c4milo has joined the channel
[00:51] scoates: I may have solved it by just keeping my http request wrapper aware of the current "depth"
[00:53] scoates: thanks again. (-:
[00:55] _announcer: Twitter: "Well, v8 and I do not node.js time chasing up." [ja] -- Yuuki Takano. http://twitter.com/ytakanoster/status/21273549925
[00:56] Tim_Smart: rauchg_: Hey! You want to talk sometime?
[00:57] softdrink has joined the channel
[00:57] twoism has joined the channel
[00:58] _announcer: Twitter: "just installed node.js - gonna see what that's all about" -- Dustin Whitney. http://twitter.com/dustinwhitney/status/21273750201
[00:58] scoates: mscdex: if you care... http://github.com/fictivekin/webshell ; now I can add a toolbox function like this, without the REPL's prompt being displayed: http://paste.roguecoders.com/p/07cd775def8dfe1e3654918427efef6b.txt
[00:59] scoates: (yes, I realize it's still complicated, and not very grokable, but ... like I said.. if you care (-: )
[01:00] mscdex: scoates: cool. good deal
[01:00] _announcer: Twitter: "@dustinwhitney I'm exploring Express.js at the moment myself - lightweight sinatra/bottle.py style web framework for node.js" -- Brendan W. McAdams. http://twitter.com/rit/status/21273880909
[01:02] scoates: I wish readline.js was a little better at handling multiline...
[01:02] scoates: or rather: lines that are longer than the current window is wide
[01:03] ezmobius has joined the channel
[01:05] foobarfighter has joined the channel
[01:06] drudge has joined the channel
[01:06] ian`` has joined the channel
[01:11] jwcooper has joined the channel
[01:12] cheriot has joined the channel
[01:15] ryah: scoates: yeah, me too
[01:15] scoates: ryah: I made some improvements to readline, already, but I'm not sure I have the skillz to fix multiline.
[01:15] scoates: ryah: http://github.com/fictivekin/webshell/blob/master/wsreadline.js
[01:16] scoates: ryah: thanks for exposing Interface, btw (-:
[01:16] jwcooper: anyone know of a server side syntax highlighter that would work with node?
[01:19] Tim_Smart: jwcooper: Well any javascript syntax highlighter really.
[01:19] Tim_Smart: As in written in Javascript.
[01:19] TangoIII has left the channel
[01:19] jwcooper: Thanks, found this one that is CommonJS...should work..http://alexgorbatchev.com/SyntaxHighlighter/whatsnew.html
[01:20] zomgbie has joined the channel
[01:23] a_meteorite has joined the channel
[01:23] confoocious has joined the channel
[01:24] confoocious has joined the channel
[01:24] _announcer: Twitter: "Just finished a Node.js/Redis prototype. Took a while to get used to all the callbacks. Upshot: handles 200+ requests/sec on 1 Heroku dyno!" -- Scott Newman. http://twitter.com/greencoder/status/21275485882
[01:27] johnster: Question for any encoding experts out there ... got a Buffer with an unsupported encoding (iso-8859-1 / latin1) and i'd like to work with it as a String - can i just loop through the bytes of the Buffer, put them through the chartable and shazzzam?
[01:30] t3h_awesome has joined the channel
[01:32] _announcer: Twitter: "Another deep-dive into NodeJS. Server-side javascript is truly an impressive thing. Love me some Google V8." -- Brian Celenza. http://twitter.com/bcelenza/status/21275980044
[01:34] a_meteorite: What's the best way to have feature requests added in? It's simple, I just rather not dabble with V8 just yet and try to patch it myself.
[01:35] charlesjolley- has joined the channel
[01:35] zapnap has joined the channel
[01:36] tekky has joined the channel
[01:38] Tim_Smart: Does connect no longer ship with a cli tool?
[01:38] matt_c: a_meteorite: http://nodejs.org/#contributing has information about contributing.  There's also the issues tracker: http://github.com/ry/node/issues
[01:38] sh1mmer: a_meteorite: what are you trying to build? It may not have to be in C
[01:39] sh1mmer: johnster: latin1 should be a subset of UTF-8
[01:39] Tim_Smart: a_meteorite: Most of Node.js is implemented in Javascript.
[01:39] a_meteorite: sh1mmer: Well, buffer.write lets you set the offset, and if the encoding is utf8, it won't write a partial utf8 character if there isn't enough room in the Buffer
[01:39] sh1mmer: johnster: you can also iterate over the bytes, but that's probably the least optimal solution
[01:40] a_meteorite: My proposal is an optional limit, so say you pre-alloc a large buffer, but you only want to write a partial string, but don't want any partial utf8 characters
[01:40] sh1mmer: a_meteorite: you could write that in JavaScript fairly easily
[01:41] a_meteorite: I took a look at the source and it appears node_buffer.cc's Buffer::Utf8Write just needs an added optional argument
[01:41] sh1mmer: a_meteorite: sending a JavaScript patch is a good way to encourage someone to rewrite it in C if needed
[01:41] brapse has joined the channel
[01:42] kriszyp_ has joined the channel
[01:42] t3h_awesome: I'd like to create an http server in node that routes GETs to function calls without having to do a giant if-elseif or switch statement. (http://server/Function1 maps to function1() http://server/Function2 maps to function2() etc.) I'd like to avoid using eval() on the URL path directly for security reasons (arbitrary code execution). Is there any way natively in node to do something like that?
[01:42] a_meteorite: Yeah, seems like a lot of work when all you have to do is accept the optional arg in that function and instead of buffer->length, it's just maxLen or something. I know if I do a patch for V8 I'll probably break it. :)
[01:43] sh1mmer: a_meteorite: well you don't have to patch, you can of course file a feature request
[01:43] sh1mmer: personally I would just take a shot if it's simple
[01:43] sh1mmer: plus it'll make it more likely to happen
[01:44] sh1mmer: and you learn something
[01:44] sh1mmer: hacking is fun ;)
[01:44] a_meteorite: Yeah, I just hate to be blamed when stuff breaks :P
[01:44] sh1mmer: a_meteorite: well if you submit a patch whoever accepts it is responsible
[01:44] sh1mmer: they should be checking your code
[01:46] Tim_Smart: In pretty much all cases, ryah will have to review it a_meteorite :)
[01:46] mscdex: get Crockford to check your code! :-D
[01:47] a_meteorite: hehe
[01:47] JimBastard has joined the channel
[01:48] a_meteorite: I guess I'll have to make myself a github account and do that code contributor agreement... :)
[01:48] Tim_Smart: Yup.
[01:48] mscdex: dude JB, where is your bot? we can't tweet without it!
[01:49] piotrekm has joined the channel
[01:50] piotrekm: Hello, is "Stream is not writable" error during write() call a bug or should I rewrite my code somehow?
[01:50] a_meteorite: Ah, mailing lists. The absolute horror.
[01:51] mscdex: piotrekm: not a bug. it means your stream isn't writable :-). check stream.writeable to make sure it == true first before calling stream.write()
[01:52] _announcer: Twitter: "Got Protovis running on a node.js+jsdom server: a step closer to beautiful data visualizations for low-end mobile devices." -- James Pearce. http://twitter.com/jamespearce/status/21277258587
[01:52] sh1mmer: piotrekm: can you gist the code?
[01:53] sh1mmer: hm protovis seems cool. Although I have no idea why if you were moving from SVN to Git you would move to Gitorious
[01:53] piotrekm: sh1mmer: it's long and a bit dirty. 
[01:54] sh1mmer: piotrekm: doesn't matter
[01:54] _announcer: Twitter: "Male who is male uses Javascript http://is.gd/ejnsU is 2008 but this time of Node.JS worth reviewing" [pt] -- lucabastos. http://twitter.com/lucabastos/status/21277395717
[01:54] sh1mmer: piotrekm: we just want to get you up and running
[01:55] piotrekm: sh1mmer: http://pastebin.com/B8u1B4BW , line 161
[01:55] mscdex: sh1mmer: well you can run your own gitorious server for free, for one thing heh
[01:55] _announcer: Twitter: "Added a bunch of stuff to Webshell this weekend. Trying to keep up with @ehaas's pace: he's got jQuery working in the envjs branch! #node.js" -- Sean Coates. http://twitter.com/coates/status/21277482767
[01:55] sh1mmer: mscdex: but if you have an open source project Github is free
[01:55] tekky: lmao @ _announcer's wonderful translation skills again
[01:55] mscdex: sh1mmer: sure, if you don't mind making your code public i guess
[01:56] sh1mmer: I hate to declare a winner, but why not hang out where everyone is?
[01:56] scoates: wow... does _announcer have a firehose feed? that was quick.
[01:56] mscdex: for keeping your git stuff local on your own servers, gitorious is about the best there is...
[01:57] mscdex: for free anyway
[01:57] tekky: gitosis is nice
[01:57] scoates: yeah, gitosis works for me
[01:57] tekky: no web interface though
[01:57] tekky: i dont think
[01:57] tekky: plus the gitosis guy is awesome at support/help and very responsive
[01:58] mscdex: i prefer to work with a web interface
[01:59] sh1mmer: I guess my point is for open source you want to encourage collaboration
[01:59] tekky: sh1mmer: true
[01:59] matt_c: scoates: there's a streaming search feed.
[01:59] sh1mmer: I could see the argument for a business
[02:01] technoweenie: for business you want things to run and have backups and shit
[02:01] dipser_ has joined the channel
[02:01] scoates: oh? nice, matt_c. new?
[02:01] mscdex: don't get me wrong, github is the best (imho) for hosting open source projects. but if you're a small business that's trying to keep costs down and don't want to reveal your commercial product's source code to the world, you need something else
[02:01] sh1mmer: piotrekm: can you gist your error too?
[02:01] matt_c: scoates: I'm not sure. http://dev.twitter.com/pages/streaming_api_methods#statuses-filter
[02:02] mscdex: one of these days i'll try to install gitorious heh
[02:02] technoweenie: yea gitosis is fine if you dont mind running your own crap.  it starts to suck if you do contracting and have lots of users/repos to manage
[02:02] pydroid has joined the channel
[02:02] sh1mmer: technoweenie: I think the argument for a business is competition is good for the market
[02:02] piotrekm: sh1mmer: http://pastebin.com/SNGqKM5x
[02:03] vladikoff_ has joined the channel
[02:03] technoweenie: competition?
[02:03] technoweenie: for hosting your source code?
[02:03] vladikoff_: hey guys i'm getting " Error: Cannot find module 'ext'" and "git submodule update --init" supposed to fix that, but how?
[02:03] nerdEd has joined the channel
[02:03] piotrekm: from documentation I assume that when stream.writeable is false, it will never get true, right? ("A boolean that is true by default, but turns false after an 'error' occurred or end() / destroy() was called.")
[02:03] sh1mmer: technoweenie: sure, having just one Git hosting company isn't a good thing
[02:04] scoates: matt_c: thanks
[02:04] technoweenie: sh1mmer: i dont know what you're talking about
[02:04] mattly has joined the channel
[02:05] mscdex: piotrekm: not necessarily. if you find that stream.write() is returning false, try waiting for the 'drain' event to occur first before writing again
[02:05] ehaas has joined the channel
[02:05] technoweenie: sh1mmer: we were talking about from the perspective of a bootstrapped startup i thought
[02:05] technoweenie: but anyways
[02:05] sh1mmer: technoweenie: it's a good thing that there are more commercial options for your Git hosting than just Github.
[02:05] technoweenie: sh1mmer: no one said it wasnt?  i surely didnt
[02:05] technoweenie: some decent competition would be nice, bring it on
[02:05] sh1mmer: technoweenie: I know. Didn't mean to imply you did.
[02:06] mscdex: piotrekm: if it's writing to a socket, check the socket's readyState to make sure writeable is not false because the connection was lost or something
[02:06] technoweenie: ha ok
[02:07] vladikoff_: what happened to 'class.js' ?
[02:07] mscdex: what is class.js?
[02:07] mscdex: heh
[02:07] blogometer has joined the channel
[02:08] mscdex: vladikoff_: what are you trying to use?
[02:08] vladikoff_: mscdex, i am trying to install hummingbird and it asked me to  "git submodule update --init" and that tried to find Clone of 'git://github.com/visionmedia/class.js.git' but that does not exist anymore
[02:10] mscdex: hmm.. i don't see that in the hummingbird repo's .gitmodules
[02:10] JimBastard: mscdex: remember javascript is object oriented 
[02:10] JimBastard: you must create classes
[02:10] JimBastard: and objects
[02:10] JimBastard: and stuff
[02:10] JimBastard: you've got no class.js
[02:10] mscdex: JimBastard: NOU
[02:11] JimBastard: im object disoriented 
[02:11] vladikoff_: eh i ll just npm the whole thing
[02:11] JimBastard: vladikoff_: npm ++
[02:11] vladikoff_: JimBastard++
[02:11] piotrekm: mscdex: thanks.
[02:12] mscdex: ++npm;
[02:12] mscdex: more efficient!
[02:12] mscdex: :-D
[02:13] JimBastard: im making cap for node, its awesome
[02:13] JimBastard: calling it noc 
[02:13] JimBastard: built on node-control
[02:13] mscdex: cap?
[02:13] JimBastard: this ruby fag thing
[02:13] mscdex: 
ACTION shakes a fist at ruby

[02:13] JimBastard: they've figured a few things out for sure
[02:13] mscdex: yeah, like how to convert people to node
[02:14] mscdex: :-D
[02:14] JimBastard: shit we got technoweenie and tmm1 :-)
[02:15] technoweenie: 'ruby fag', seriously?
[02:15] JimBastard: technoweenie: i couldnt figure out how to spell capistrono 
[02:15] JimBastard: >.<
[02:15] JimBastard: http://www.capify.org/index.php/Capistrano
[02:15] JimBastard: there we go
[02:17] technoweenie: this community isn't as welcoming as ruby though
[02:17] technoweenie: its more like 4chan
[02:18] Blink7 has joined the channel
[02:18] _announcer: Twitter: "node.js may be fast, but its community needs to be more welcoming and less like 4chan" -- technowürst. http://twitter.com/technoweenie/status/21279034153
[02:18] scoates: I've found channel, in particular, to be quite helpful, even when I ask potentially very stupid questions
[02:18] piotrekm: ok. And is there any way to provoke stream not to be writeable? It's quite hard to test
[02:19] technoweenie: scoates: i've been cursed out for asking dumb questions
[02:19] JimBastard: i dunno, almost every single "rails developer" ive met is arrogant. almost every single "javascript developer" i've met is humble from getting his ass kicked by the browser SOO MANY TIMES
[02:19] sh1mmer: technoweenie: 'ruby fag'?
[02:20] daleharvey: to be honest the node.js community is something I would like to try and replicate with erlang, its a pretty great community
[02:20] JimBastard: i try to help people when im not yelling at them :-(
[02:21] bryanl has joined the channel
[02:21] daleharvey: erlang community isnt the strongest when it comes to the web side of things
[02:21] technoweenie: well, some bad apples can ruin the experience for everyone
[02:21] blogometer: I'm happy with the Node.js community.
[02:21] sh1mmer: technoweenie: yeah, really.
[02:22] blogometer: I believe it will get worse as more people show up, though.
[02:22] sh1mmer: blogometer: I'd still like to see a high standard now.
[02:22] JimBastard: for sure, its gonna be php 10x
[02:22] technoweenie: well then you start having sub communities
[02:22] blogometer: Rails fell into software bureaucracy a while back.
[02:23] blogometer: How man HTTP verbs can dance on the head of pin.
[02:23] JimBastard: sh1mmer and technoweenie, #node.js.is.serious.business 
[02:23] twoism has joined the channel
[02:23] JimBastard: see you later! :-)
[02:24] daleharvey has joined the channel
[02:24] sh1mmer: man, I can't believe it took me this long to really give Vi a chance
[02:24] Aria: I think something that might help the community is having a separate, curated standard library -- not a pile of packages in a package manager, but something you can install coherently.
[02:24] sh1mmer: Aria: I think we'll get there
[02:24] sh1mmer: but first we need to find winners
[02:25] sh1mmer: right now is the gold rush
[02:25] jb55: something like the haskell platform?
[02:25] Aria: Yeah. But I think the time to start is now -- and take Ryah's approach: Break things, but as little as possible.
[02:25] sh1mmer: while people a) figure out what works b) people start to show leadership over various peices
[02:25] JimBastard: as projects get real stakeholders....
[02:25] JimBastard: 
ACTION turns on the money machine

[02:26] blogometer: I like Node.js because it is about as complicated as C.
[02:26] Aria: Collect. Curate. Make a set of tools that work together.
[02:26] blogometer: I don't want APIs.
[02:26] blogometer: And I prefer CPAN to Boost.
[02:26] vladikoff_: where does npm store all the fun stuff?
[02:27] sh1mmer: you mean http://npm.mape.me/
[02:27] Aria: CPAN has its issues. It's really hard to deploy and doesn't always work together.
[02:28] blogometer: The C++ STL is an example of a curated standard library.
[02:28] blogometer: Right. Which makes it awesome.
[02:29] SubStack: s/esome/ful/
[02:29] blogometer: It is an ecosystem that is constantly evolving.
[02:29] SubStack: the stl is an ecosystem?
[02:29] blogometer: STL is awful. 
[02:29] blogometer: CPAN is awesome.
[02:29] SubStack: ah that
[02:30] SubStack: yeah, I wish more of cpan came bundled in the standard dist
[02:30] blogometer: Node.js is incredibly concise.
[02:30] vladikoff_: "Error: Cannot find module 'old'" wtf is that even 
[02:30] blogometer: I don't want to bring in SOUNDEX if I'm not going to use it.
[02:30] aaron___ has joined the channel
[02:31] tekky: vladikoff_: old, outdated, not new or current? :P
[02:31] ddollar has joined the channel
[02:31] blogometer: Or BerkleyDB. 
[02:31] vladikoff_: tekky, someone called a module 'old'
[02:32] blogometer: Does anyone have underscore.js in an NPM?
[02:34] JimBastard: blogometer: npm list
[02:34] JimBastard: underscore@1.0.3                         =documentcloud remote underscore@1.0.4                         =documentcloud latest stable remote 
[02:34] JimBastard: npm install underscore
[02:34] blogometer: Nice!
[02:34] blogometer: Thank you.
[02:34] _announcer: Twitter: "Any recommendations for "getting started with node.js" screencasts? Hoping to get up to speed while I have my hands full with the baby." -- Elliott Draper. http://twitter.com/ejdraper/status/21280102374
[02:41] aaron___: ryah: you there? :)
[02:41] cataska has joined the channel
[02:43] aconran__ has joined the channel
[02:47] tilgovi_ has joined the channel
[02:48] ajpiano has joined the channel
[02:49] tilgovi_ has joined the channel
[02:50] _announcer: Twitter: "http://nodejs.org/ &lt;-- This is extremely relevant to my interests.  #awesome #notjquery" -- shadowmint. http://twitter.com/shadowmint/status/21281081430
[02:52] jimt has joined the channel
[02:53] _announcer: Twitter: "@technoweenie got that same feeling when I saw a node.js user posting that ruby was the new cobol. painfully stupid." -- Christopher Burnett. http://twitter.com/twoism/status/21281318247
[02:56] jakehow has joined the channel
[02:57] brapse has joined the channel
[02:59] astrolin has joined the channel
[03:00] _announcer: Twitter: "playing around http://nodejs.org/" -- Roc Yu. http://twitter.com/uiicor/status/21281774656
[03:00] jimt_ has joined the channel
[03:01] mr_danie1 has joined the channel
[03:05] kodisha has joined the channel
[03:07] jamescarr has joined the channel
[03:07] jamescarr: that zalgo stuff is... odd
[03:09] jimt has joined the channel
[03:12] joshbuddy has joined the channel
[03:12] blogometer: So...
[03:12] blogometer: Node.js depends on libeio for great justice.
[03:12] blogometer: And libeio is a creation of one Marc Lehmann.
[03:13] aaron___: process.uncaughtException for the win!
[03:14] aaron___: (so enthused, had to share it somewhere)
[03:17] charlesjolley- has joined the channel
[03:18] jamescarr: libeio?
[03:18] jamescarr: never saw that?
[03:18] pheuter has joined the channel
[03:20] mattikus_ has joined the channel
[03:24] vladikoff_ has left the channel
[03:26] mattly has joined the channel
[03:28] admc has joined the channel
[03:30] justin_ has joined the channel
[03:32] jamescarr: argh... is there someway to get the visitor's IP address in expressjs? I am sure there is since it gets logged to the console
[03:32] jamescarr: just dont see it
[03:34] pheuter has joined the channel
[03:35] jimt has joined the channel
[03:42] mjr_: technoweenie: are we really not welcoming?  That's rather surprising to me.
[03:44] jamescarr: nvm figured it out
[03:45] _announcer: Twitter: "Countdown to Knockout: Post 1 - How to Install Node.js http://bit.ly/cOU2AM" -- node knockout. http://twitter.com/node_knockout/status/21284709426
[03:46] jimt_ has joined the channel
[03:51] jimt has joined the channel
[03:53] _announcer: Twitter: "@voodootikigod @technoweenie how is Node.js’s community like 4chan? o_O" -- ell.io. http://twitter.com/elliottcable/status/21285196065
[03:53] elliottcable: technoweenie 3↼ what do you mean about the 4chan thing? (cf. tweet)
[03:54] PyroPeter has joined the channel
[03:54] _announcer: Twitter: "mention for my #nodejs patches in http://dailyjs.com/2010/08/13/ntpl/" -- trentmick. http://twitter.com/trentmick/status/21285288709
[04:01] benburkert has joined the channel
[04:02] jimt_ has joined the channel
[04:04] matt_c: elliottcable: http://nodejs.debuggable.com/2010-08-16.txt @ 02:17.
[04:06] dcolish has left the channel
[04:08] mattikus has joined the channel
[04:08] overra has joined the channel
[04:16] twoism has joined the channel
[04:23] derferman has joined the channel
[04:24] elliottcable: matt_c 3↼ looks like mostly JimBastard being 4chan-ish
[04:24] elliottcable: technoweenie 3↼ so, yeah, I’m still lost as to how the *whole community* feels like 4chan. /-:
[04:24] Draggor has joined the channel
[04:25] elliottcable: technoweenie 3↼ in fact, I’ve found this community several orders of magnitude more welcoming than that of Ruby (note: I’m banned in two high-profile Ruby channels.)
[04:25] elliottcable: technoweenie 3↼ so, yes, I’m quite interested in what, exactly, you find so unwelcoming, or ‘4chan-ish,’ so to speak around here :3
[04:25] Tim_Smart: And ##C - it seems they are all are far superior to other human beings.
[04:26] elliottcable: yep
[04:26] elliottcable: well, I’m used to ##C.
[04:26] elliottcable: Actually, most of ##C are fine; it’s just Zhivago being his usual, assish, self. If you don’t let him get to you, it can actually be quite the hoot :3
[04:26] Aria: Heh, yeah. Actually, I find the same about #ruby-lang mostly. 
[04:27] Tim_Smart: elliottcable: Yeah I found it rather humerous.
[04:28] Tim_Smart: *humorous 
[04:28] mape: elliottcable: Believe it was statements like "this ruby fag thing" in this channel that made him feel like that
[04:28] tpryme has joined the channel
[04:28] Tim_Smart: elliottcable: And twzm's poetry was... interesting.
[04:29] Tim_Smart: *twkm
[04:29] mikeal has joined the channel
[04:30] hij1nx has joined the channel
[04:31] keeto has joined the channel
[04:35] keeto_ has joined the channel
[04:36] Tim_Smart: mape: Didn't you have #node.js stats page somewhere?
[04:36] mape: http://stats.nodejs.se/
[04:37] saikat has joined the channel
[04:37] Tim_Smart: Haha that sums it up reasonably well. JimBastard talks the most, yet gets the least amount of 'thanks'
[04:38] keeto_ has joined the channel
[04:40] jimt has joined the channel
[04:41] benburkert has joined the channel
[04:44] jakehow has joined the channel
[04:46] technoweenie: ungrateful bastards
[04:47] keeto has joined the channel
[04:53] stride: technoweenie: http://en.wiktionary.org/wiki/irony don't see how one statement by JimBastard makes the comunity 4chan'y
[04:53] jsilver has joined the channel
[04:54] technoweenie: thats just the most recent one
[04:56] stride: the channel helped many people, most recently this stephen guy, to get started, don't see anything 4chan'y there you don't have everywhere in irc, sorry
[04:57] jesusabdullah: It's a good channel
[04:57] technoweenie: well, i've been cursed out for asking a newb question
[04:57] technoweenie: but yea, things are fine, keep on doing what you're doing
[04:57] jesusabdullah: Really?
[04:57] jesusabdullah: That's surprising
[04:57] jesusabdullah: Here, right?
[04:58] Tim_Smart: Every community has a ratio of helpful people:not-so-helpful people
[04:59] Tim_Smart: I think node is at a pretty healthy 50:1
[04:59] Tim_Smart: or something :D
[05:00] jsilver has joined the channel
[05:01] paulwe has joined the channel
[05:01] jsilver has joined the channel
[05:02] _announcer: Twitter: "lisp.js: lisp on node. http://ajxn.it/aesyMY // personally, I already have my lisp on node... and it is JavaScript!" -- Dion Almaer. http://twitter.com/dalmaer/status/21289255230
[05:03] technoweenie: yea its fine... when communities get too big they splinter off anyway.  most of my irc channels are pretty focused communities
[05:04] elbart has joined the channel
[05:08] technoweenie has left the channel
[05:08] falconair has joined the channel
[05:10] mattly has joined the channel
[05:10] falconair: is anyone here familiar with something called the "FIX protocol?"
[05:11] Aria: http://www.fixprotocol.org/what-is-fix.shtml .. Only trivially aware of it. I ran into it a year or so ago when I was looking for information on OFX
[05:11] Aria: Looks like the usual nightmare of industry consortium spec.
[05:12] falconair: hm...I have a half-assed, incomplete implementation of it in node, if someone knows the protocol, i wanted to get some ideas about code organization
[05:17] ryankshaw has joined the channel
[05:20] paulwe has joined the channel
[05:24] paulwe has joined the channel
[05:24] jackish has joined the channel
[05:39] paulwe has joined the channel
[05:40] Draggor1 has joined the channel
[05:40] shreekavi has joined the channel
[05:45] jsilver has joined the channel
[05:46] jetienne has joined the channel
[05:47] _announcer: Twitter: "[Js 人 気 情报] lisp.js: a Lisp for node.js - lisp-js&#39;s posterous http://bit.ly/cLV6n0 # javascript" [de] -- jsMagazine. http://twitter.com/jsMagazine/status/21291623383
[05:48] joshbuddy has joined the channel
[05:51] jochen has joined the channel
[05:52] _announcer: Twitter: "npm config chain: C=cli, E=env, U=userconfig, G=globalconfig, D=default  http://vimeo.com/14120299  #nodejs" -- trentmick. http://twitter.com/trentmick/status/21291872034
[05:53] brapse has joined the channel
[05:54] pdelgallego has joined the channel
[05:55] _announcer: Twitter: "I have seen many people playing with Node.Js, I think then I&#39;ll curiara it: D" [pt] -- Charleno Pires. http://twitter.com/charlenopires/status/21292005197
[05:57] amerine has joined the channel
[05:58] sYsk has joined the channel
[05:58] sYskk has joined the channel
[06:01] saikat: rauchg_: i remembering you mentioned something about client-side disconnect notifications not being reliable.  about how reliable are they?  or do should i build in a timeout in my app?
[06:01] saikat: i remember*
[06:01] SamuraiJack has joined the channel
[06:01] rauchg_: saikat: essentially 
[06:01] rauchg_: the client and server have to both share the same heartbeat timeout 
[06:01] rauchg_: if the client hasn't received a heartbeat past that timeout 
[06:01] rauchg_: it considers it disconnected
[06:02] rauchg_: for certain transports, we might know sooner 
[06:02] rauchg_: (eg: websocket onclose)
[06:02] rauchg_: some of these `sooner` scenarios are handled 
[06:02] rauchg_: that's why it's incomplete 
[06:02] rauchg_: i'm probably gonna work on it this week
[06:02] saikat: hm... but you're saying though that
[06:02] saikat: if a disconnect happens
[06:02] saikat: i will know - just in some cases i might not know immediately
[06:02] rauchg_: no, sometimes you'll never know :D
[06:02] saikat: oh fair
[06:03] rauchg_: because i didn't implement the client-side timeout 
[06:03] rauchg_: granted
[06:03] rauchg_: those cases are really rare 
[06:03] rauchg_: and what's more important is the server side disconnection notification 
[06:03] saikat: yeah
[06:03] saikat: definitely true
[06:03] rauchg_: because of cleanup tasks, tcp allocation, memory freeing tasks etc
[06:03] rauchg_: but 0.6 will definitely have it 
[06:04] saikat: a timeout on connect would be nice too - tohugh i just put one in for my purposes and it's of course not terribly hard
[06:04] rauchg_: indeed.
[06:05] rauchg_: however
[06:05] rauchg_: we don't want to do much more than WebSocket
[06:05] saikat: websocket has a connection timeout?
[06:05] rauchg_: but arguably, websocket is not enough as it is for most applications
[06:05] rauchg_: nop
[06:05] saikat: oh you are saying we don't want to add too much on top
[06:05] rauchg_: exactly
[06:05] saikat: that's fair
[06:05] rauchg_: there's a good article
[06:05] rauchg_: that illustrates how much more you have to do on top of websocket
[06:05] rauchg_: http://cometdaily.com/2010/03/02/is-websocket-chat-simple/
[06:06] rauchg_: for example, onclose="" isn't reliable either 
[06:06] rauchg_: and nor is waiting for the FYN on the server side
[06:06] rauchg_: FIN*
[06:07] hij1nx has left the channel
[06:07] saikat: i see
[06:07] hij1nx has joined the channel
[06:09] ph^ has joined the channel
[06:09] rauchg_: as things stand today, we're already doing more than WebSocket.
[06:09] rauchg_: because of the inherent limitations of some of the transports and the fact that we share the same protocol across all of them 
[06:09] rauchg_: and ryah wants fireAndForget
[06:09] rauchg_: :P
[06:09] _announcer: Twitter: "async.js for node.js now with file globbing - ls: async.glob("*.txt").get("path").print().end() http://tinyurl.com/3afna2r" -- Fabian Jakobs. http://twitter.com/fjakobs/status/21292738542
[06:10] saikat: right
[06:10] saikat: but that's why i like socket.io and don't just use websockets =)
[06:10] rauchg_: yep 
[06:11] saikat: also, one other question - is socket.connected keeping track of connected the way i think it is?
[06:11] rauchg_: but we don't want to go over too much
[06:11] saikat: definitely
[06:11] saikat: i mean - it might make sense to make a layer on top of socket.io but
[06:11] saikat: i do think a layer somewhere should exist - cometd sounds nice
[06:12] rauchg_: for now you're better of setting your own connected flags in a closure
[06:12] saikat: ok
[06:12] rauchg_: because iirc im setting connected to false when XHR GET finishes
[06:12] rauchg_: then to true when the client connects again
[06:12] rauchg_: (in polling)
[06:12] saikat: ah i see
[06:12] rauchg_: that should probably be an internal one, and we should have a public .connected one 
[06:12] rauchg_: that is the `absolute` one
[06:13] saikat: ok - thanks as usual
[06:15] rauchg_: np
[06:17] sYskk: question here: how should I handle processor heavy tasks with node.js without blocking the http server?
[06:17] Throlkim has joined the channel
[06:18] Aria: start another process. Or break it up into pieces.
[06:18] sYskk: ok, how to start a new process?
[06:18] mjr_: sYskk: you can use the web workers module for that.
[06:18] Aria: (... Can your computer actually handle the workload? Or is it just faking it?)
[06:18] tango3 has joined the channel
[06:18] mjr_: or you can just spawn child processes
[06:20] sYskk: right
[06:21] pquerna: or you could limit the number of incoming http connections
[06:21] pquerna: 
ACTION retreats back into the realm of limits

[06:22] mjr_: pquerna: don't forget to come back to the world of rewriting SSL.
[06:22] mjr_: we need you there!
[06:22] pquerna: mjr_: http://github.com/pquerna/selene
[06:22] pquerna: mjr_: its not done. no promises.
[06:22] mjr_: awesome
[06:22] felixge has joined the channel
[06:22] felixge has joined the channel
[06:22] pquerna: i'll try to work on it next weekend again.
[06:23] mjr_: super cool
[06:23] mjr_: I'm glad you are working on it.
[06:26] persson has joined the channel
[06:26] felixge has joined the channel
[06:26] felixge has joined the channel
[06:27] astrolin_ has joined the channel
[06:31] joshbuddy_ has joined the channel
[06:34] Egbert9e9 has joined the channel
[06:34] sYskk: is there any code profiling tool for node.js?
[06:35] mjr_: sYskk: you can use the V8 profiler for many things.
[06:35] mjr_: But it requires that you build d8, which is sometimes hard.
[06:35] mjr_: If you get it all working though, the V8 profiler has some good information in there.
[06:36] jackish_ has joined the channel
[06:41] terinjokes: mjr_: your response above made me realize how similar the "B" and "8" look in this typeset
[06:42] mjr_: 8oy, that must 8e confusing.
[06:42] EyePulp: heh
[06:43] rauchg_: sln_brigade_destroy()
[06:44] terinjokes: mjr_: well, now that'm aware, I know that the 8 has a slight hint of a curve in the lower bottom corner (and, thus, know that you used an eight in your reply)
[06:45] jackish has joined the channel
[06:45] mjr_: busted!
[06:45] elliottcable: mape 3↼ hahaha, “ruby fag thing?” what?
[06:45] terinjokes: i guess outside of sourcecode, I can identify based on context anyways :/
[06:50] mape: elliottcable: what Jim said
[06:55] Dmitry1 has joined the channel
[07:02] tisba has joined the channel
[07:05] elliottcable: mape 3↼ well, that‘s Jim. What can you expect?
[07:06] keeto has joined the channel
[07:06] mape: Well, if that makes other people disslike the community it isn't really about what I think about it :)
[07:07] luddep has joined the channel
[07:10] javajunky has joined the channel
[07:14] Blink7 has joined the channel
[07:24] stride: meh, he ranted a bit, so what?
[07:24] micheil has joined the channel
[07:27] tango3 has joined the channel
[07:36] tekky has joined the channel
[07:36] sYskk has joined the channel
[07:37] sYskk: i have a mental block with node.js. i've started tinkering with it and even considered building my web app with it
[07:37] sYskk: however there's something thats bugging me
[07:37] sYskk: im scared to death of writing soome blocking code without realizing it
[07:37] alcuadrado: in fact, you do write blcking code most of the time, even in node :p
[07:38] alcuadrado: "1 + 1" is blocking
[07:38] sYskk: yea but i mean blocking code in the range of 1 second
[07:38] alcuadrado: just care about IO, networking, user input, and else
[07:38] sYskk: that would mean i could lose some http requests right?
[07:39] sYskk: maybe im over panicking :P
[07:39] alcuadrado: I don't think so... they would just be slowler (wating)
[07:39] alcuadrado: but that should be asked to someone more versed in node, as ryah :p
[07:39] virtuo has joined the channel
[07:40] alcuadrado: but AFAIK, there is no blocking module available in the default node distribution
[07:41] stride: your OS is queueing new connections to a port, but if you really have blocks that really take up a full second of processing I guess you should take a look at that webworker module or something
[07:41] _announcer: Twitter: "@cramforce Or node.js ftm— could be a preprocessor build into the runtime, too." -- Thomas Fuchs. http://twitter.com/thomasfuchs/status/21296963669
[07:45] sYsk has joined the channel
[07:45] mjr_: sYskk: Your OS has a "backlog" of unaccepted TCP connections that it uses for this sort of thing.  
[07:45] jblanche has joined the channel
[07:46] mjr_: But for most things, node can handle thousands of requests per second, so you'd have to do quite a bit of work in general to make it drop new connections.
[07:46] mjr_: I say just go for it, and if it turns out what youa re doing is too slow, deal with that then.
[07:47] sYsk: ok
[07:48] terinjokes: and *that* my friends is why I keep my electronic equipment on surge protectors with battery backup. The client computer didn't fair that surge very well
[07:48] micheil: morning' folks
[07:49] ditesh|cassini has joined the channel
[07:49] micheil: sYsk: it's pretty damn hard to write blocking code in node.js, unless you explicitly call a blocking method.
[07:50] sYsk: lets say i need to regex some user input to eliminate some stuff, wouldnt that be blocking for a while?
[07:50] micheil: alcuadrado: also, "1+1" isn't really blocking per-say, it's procedural, and the time it takes to calculate 1+1 should be incredibly low, otherwise you have bigger problems then say blocking file i/o
[07:50] micheil: sYsk: type `node` into your shell, and hit enter. Try it.
[07:51] micheil: if it isn't instantaneous, then it's probably blocking
[07:51] ryah: sYsk: yeah - that ends up not being the common thing
[07:51] micheil: for things like regex, and standard operations, you're looking at < 1ms in most cases, generally always < 10ms.
[07:52] sYsk: which means every request will be frozen until the regex finishes right?
[07:52] sYsk: ok
[07:52] micheil: sYsk: seriously, try it. If you can write a regex that takes longer then 50 ms, I'd be surprised.
[07:52] sYsk: that thought kinds of scares me :)
[07:52] ryah: sYsk: the usual answer is to split of the regex parsing into chunks
[07:53] ryah: s/parsing/execution/
[07:53] ryah: don't do a 10mb file in one go, that is
[07:53] micheil: ryah: but.. but.. I like my over complicated regular expressions that few other people can sanely understand :(
[07:53] micheil: ;P
[07:53] exlibris has joined the channel
[07:53] alcuadrado: micheil, I know 1+1 is made in a single CPU instruction, but what I mean, is that no of the following statements would be executed in the meanwhile (althought it might be something like 1 / cpu frecuency seconds, i think :P)
[07:53] micheil: ryah: probably next wednesday I'll start on reworking the documentation. I'll create a branch for it.
[07:54] derferman has joined the channel
[07:54] micheil: alcuadrado: something, somewhere along the line has to block, theoretically, I'm sure.
[07:54] micheil: the idea is to have non-blocking where it's actually going to take several thousand / million / trillion cpu cycles later
[07:55] alcuadrado: yeah, I think so... and if not, that would be a pain in the ass though
[07:55] micheil: I think ryah has some good slides on this, don't you?
[07:55] micheil: generally CPU's are pretty damn fast, especially on a dedicated or shared server.
[07:56] micheil: where you probably have 2 if not 8 cpu's to yourself, where if something else is owning your system, then node should still be able to function properly
[07:57] mpoz2 has joined the channel
[07:57] alcuadrado: yeah, I know =)
[07:58] zomgbie has joined the channel
[07:58] stride: oh, on that matter, does blocked within node.js code mean a single function() or a complete stack of calls without using a callback?
[07:58] alcuadrado: I just wanted to show sYsk that he doesn't need to warry that much... but maybe my way wasn't the best :p
[07:58] micheil: ryah: also, would it be an idea to make the event emitter emit in a non-blocking way?
[07:59] sYsk: im worried because this is a totally new paradigm for me
[07:59] _announcer: Twitter: "@kuvos I think this is actually more interesting for backend stuff, node.js for example." -- Thomas Fuchs. http://twitter.com/thomasfuchs/status/21297720974
[07:59] micheil: for example, in my Evan.js project, which is extremely similar to event emitter, I use setTimeout(callback, 0);
[07:59] sYsk: i've always worked with cgi like
[07:59] alcuadrado: sYsk, haven't you coded client-side javascript?
[07:59] micheil: sYsk: well, hopefully you'll find it more pleasurable to work with
[08:00] sYsk: I have but nothing too complex yet, mostly UI stuff
[08:00] sYsk: also used Jaxer 1 year ago to do some scraping using the dom
[08:00] alcuadrado: well, UI stuff is completetly event driven :p
[08:00] micheil: sYsk: I've got some slides you may want to check out; even though they are really introductory
[08:00] micheil: http://www.slideshare.net/simon/evented-io-based-web-servers-explained-using-bunnies
[08:00] sYsk: alcuadrado: but I'm not afraid of slowing down 1 user... the thing I'm afraid of is to have 1 user block everyone else :D
[08:01] sYsk: if I have some slow code for one particular request, it will slow down some users who do the request but at least i can fix it later without worrying that everyone will be slowed down
[08:02] alcuadrado: yeah, that's true
[08:03] alcuadrado: but I think a simple benchmark in maybe-too-slow bits of code would help
[08:03] sYsk: i.e.: http://host/very/slow/request http://host/very/fast/request ... with php/ruby/python only the people going to the slow url will be slowed down. with node.js, everyone is slowed down
[08:03] stride: yeah but that would be true for every system, no matter how the stuff for that one user takes over your cpu
[08:03] sYsk: perhaps im misunderstanding or worrying too much
[08:03] ryah: micheil: new docs - okay - need to talk abou tthat
[08:03] ryah: gtg
[08:03] micheil: ryah: how would you like to talk?
[08:04] micheil: email me, micheil@brandedcode.com
[08:04] overra has joined the channel
[08:04] micheil: (or I'll email you in a days time or something)
[08:05] stride: sYsk: php/ruby/python don't grow more cpus on your server so the http://host/very/fast/request would still at least be slowed down until very/slow/request finished stuff
[08:05] pdelgallego has joined the channel
[08:06] rnewson has joined the channel
[08:07] sYsk: slowed down but not completely blocked. to illustrate if the fast request takes 10s, maybe the slow request would take 5s instead of 1s. with node.js, the fast request would be blocked until slow request is done which means at least 10s
[08:08] sYsk: right?
[08:09] sh1mmer: sYsk: I think you are missing the way it works
[08:09] sYsk: very possible ;)
[08:10] sh1mmer: sYsk: can I give you a completely non-technical analogy?
[08:10] sYsk: go ahead
[08:10] sh1mmer: imagine a bar tender
[08:10] sh1mmer: you and some friends go to a bar and order drinks
[08:10] sh1mmer: it's a cocktail bar so some of the drinks are complex, some are simple
[08:11] sh1mmer: one guy orders a beer, the bar tender pulls it out of the fridge pops the top and hands it over
[08:11] sYsk: yes
[08:11] stride: "node.js - The drinking game." now on scribd :>
[08:11] sh1mmer: someone else ordered a smooth margarita so he mixes it puts in the blender 
[08:12] sh1mmer: and then while the blender is going he starts making another drink
[08:12] sh1mmer: until the blender finishes
[08:12] sh1mmer: then he finishes her margirita and serves it
[08:12] sh1mmer: and carries on
[08:12] sh1mmer: that's non-blocking
[08:12] sh1mmer: imagine another bar tender, who only does 1 drink at a time
[08:12] sh1mmer: he can serve beers all day long because they have a fixed time to deliver
[08:13] sh1mmer: but complex drinks take him much longer because he can't do anything while he is waiting for the blender.
[08:13] sh1mmer: and while he is waiting so is everyone else wanting a drink
[08:13] sYsk: right I understand that part
[08:13] felixge has joined the channel
[08:14] sh1mmer: So what don't you get?
[08:14] sYsk: but wouldnt a better analogy be: many bar tenders who work linearly but that are each slower than the async bar tender?
[08:14] sh1mmer: Actually I think the bit you are missing is this
[08:15] sh1mmer: In node most things are async
[08:15] alcuadrado: sYsk, thats like php, for example, works
[08:15] sYsk: right
[08:15] sh1mmer: so it's more like the bartender is an octopus
[08:15] sh1mmer: because he's quickly switching contexts doing lots of small pieces for each customer
[08:15] ntelford has joined the channel
[08:15] guybrush: you can do child-processes though, isnt it?
[08:15] alcuadrado: as you may know, v8 is far faster than php, so, if you are not conserned to, lets say, a php regex validaton, you shouldn't be with node
[08:16] guybrush: each tentakel of the octupus is a child-process :p
[08:16] sh1mmer: sYsk: the point being that a single request isn't 1 big 10 second action, it's most likely dozens of non-blocking actions
[08:17] sYsk: right
[08:17] sh1mmer: so you can squeeze a bunch of small requests in while waiting for a slow request
[08:17] sh1mmer: and the more you can split your code up to be non-blocking the better
[08:18] sYsk: is there an easy way in JS to write some non blocking code i.e.   do_this_in_parallel(function() {/* CPU heavy code */}, callback);
[08:18] sYsk: that would really end my tormenting
[08:18] sYsk: haha
[08:18] alcuadrado: here you it is in example... you may have heared about this fucking damn regex to validate every possible variation of an email address made by the perl guys http://www.ex-parrot.com/pdw/Mail-RFC822-Address.html.  Try it in node: http://www.pastie.org/1095137   If you are doing something THAT complex in a single statement, well... maybe node is not the problem :P
[08:18] sh1mmer: micheil: I have a function like var foo = function(input) { a(); b(); c(); } and b() just got refactored to be a non-blocking call
[08:18] alcuadrado: here is an example*
[08:19] sh1mmer: micheil: so I'm trying to codify a pattern for how to refactor the function foo in a useful way
[08:19] micheil: one sec sh1mmer; trying to write a slight proposal to hybi
[08:19] guybrush: sYsk: if its on the client-side you can do stuff in parallel with web workers http://wearehugh.com/public/2010/08/html5-web-workers/ , on the server side its child-processes
[08:19] sh1mmer: sYsk: you can use child processes, people have suggested slaved node processes using web workers
[08:20] sh1mmer: I should just let the mighty pirate talk
[08:20] guybrush: i dont fear LeChuck !
[08:20] sh1mmer: although I'm also going to link http://developer.yahoo.net/blog/archives/2010/07/multicore_http_server_with_nodejs.html because Peter wrote the web worker support
[08:21] stride: sYsk: the step library lets you handle parallel stuff as well
[08:21] stride: (iirc)
[08:22] sh1mmer: bleh
[08:22] sh1mmer: don't use step
[08:22] sh1mmer: it's cool Tim wrote it but just learn to do it right
[08:22] stride: hm? why not?
[08:23] stride: ah, that, yeah point taken
[08:23] sh1mmer: It's a crutch
[08:23] _announcer: Twitter: "@cramforce not that node.js is slow, of course — this is for more extreme cases :)" -- Thomas Fuchs. http://twitter.com/thomasfuchs/status/21298765456
[08:23] sYsk: I see
[08:23] stride: but it seems nice if you have seperate components and need to implement a barrier for it
[08:23] zomgbie has joined the channel
[08:23] sh1mmer: Fuchs is a menice.
[08:24] _announcer: Twitter: "@garbrand our #AmsterdamJS talks are getting more and more #nodejs packed each time... http://bit.ly/clthYV /cc @mennovanslooten" -- mikedeboer. http://twitter.com/mikedeboer/status/21298789358
[08:24] sYsk: it would be nice to be able to do it "inline"
[08:24] sh1mmer: ugh can't spell this morning.
[08:25] _announcer: Twitter: "Interesting weekend dabbling with nodejs and sproutcore; why are we still using this PHP and python crap?" -- Carl. http://twitter.com/carlmw/status/21298829798
[08:25] guybrush: php is _not_ crap
[08:26] alcuadrado: I disagree
[08:26] alcuadrado: php is the coolest crap out there =)
[08:26] alcuadrado: hehe
[08:26] _announcer: Twitter: "@humanzz have a look at node.js, sounds interesting but i still feel lost somehow understanding what is going on" -- Osama Brekaa. http://twitter.com/BioNuc/status/21298899085
[08:26] _announcer: Twitter: "I just realized that I hardly have to do any debugging on my node projects that use test driven development. #tdd #nodejs #ftw" -- Felix Geisendörfer. http://twitter.com/felixge/status/21298907960
[08:26] alcuadrado: it has many creppy things, but its pretty cool although
[08:27] felixge: hey, anybody in here interested in trying my mysql client lib?
[08:27] micheil: k, sh1mmer, back.
[08:27] felixge: http://github.com/felixge/node-mysql
[08:27] N` has joined the channel
[08:27] guybrush: php is so nice, it enabled so very much people to make awesome apps
[08:27] felixge: would love to get some feedback on the API and stuff
[08:27] micheil: felixge: I can't, I don't actually have mysql installed on my macbook pro or my webserver(s), as much as I'd love to help.
[08:28] felixge: micheil: np :)
[08:28] alcuadrado: guybrush, I work as a php programmer... but I'm actualy falling in love with JS :P
[08:28] felixge: micheil: I guess I'll have to write a node.js mysql server next
[08:28] felixge: just kidding :)
[08:28] sh1mmer: felixge: is it the world's first non-blocking mysql driver?
[08:28] felixge: sh1mmer: no, but probably the world's most decent mysql driver
[08:28] sh1mmer: heh
[08:28] stride: *git clone*
[08:28] felixge: as in readable code and non-buffering parsing
[08:28] digitalspaghetti: felixge: i'll be having a go, I have a MySQL backend that I can test on
[08:29] felixge: I just got result set parsing working last night, but all the basic stuff should be working
[08:29] felixge: sh1mmer: I looked at some of the other mysql client implementations in other languages and it got me quite scared for those big sites that actually use them :)
[08:30] sh1mmer: heh
[08:30] lachlanhardy has joined the channel
[08:30] alcuadrado: felixge, how much time do you think it remains for your driver to be ready? having mysql in node will be pretty cool (:
[08:30] ewdafa has joined the channel
[08:31] stride: any pointers on how to run the testcases in there?
[08:31] felixge: alcuadrado: I hope to start using it in production this week or next
[08:31] felixge: stride: make test
[08:31] felixge: alcuadrado: There will be a few features I have yet to add (like prepared statements), but once I'm happy with charsets and basic queries it should be quite stable
[08:32] sh1mmer: fuck
[08:32] alcuadrado: great news, I'll finish a project this week and start a node one after that, and I'd love to use it
[08:32] tekky has joined the channel
[08:32] sh1mmer: that reminds me I really have to finish input filters
[08:32] micheil: felixge: link me to the api, if you will, I'll tell you if it's awesome and understandable. (I may even changelog it.)
[08:32] felixge: I'll announce the project later on the mailing list, hopefully I can get a few more companies help sponsoring the efford, it's quite a lot of work :)
[08:32] felixge: micheil: http://github.com/felixge/node-mysql has API docs
[08:32] alcuadrado: felixge, yeah, that is not a big issue... prepeared statements can be added to the project later :p
[08:32] sh1mmer: felixge: you should hit up pgriess I wouldn't be surprised if he had a use case
[08:32] micheil: sh1mmer: step does have it's uses, and one of those is chaining callbacks for a more readable code style 
[08:33] micheil: sh1mmer: peter probably does, by the sounds of things
[08:33] felixge: sh1mmer: already emailed him, his team doesn't use mysql and the teams at yahoo that do don't work with node yet
[08:33] sh1mmer: felixge: I know.
[08:33] sh1mmer: I'm surprised he doesn't have any mySql though
[08:33] micheil: sh1mmer: mail.. doesn't use mysql? actually, they'd use hadoop, righT?
[08:33] micheil: or just smtp/imap
[08:33] felixge: haven't heard back from EXT either, they are investing in node pretty heavily so they would benefit
[08:33] Draggor1 has joined the channel
[08:33] sh1mmer: I guess given his project that makes sense
[08:34] felixge: anyway, this driver will hopefully help bigtime with node adoption. I mean all the cool kids play with NoSQL, but most of the big boys are all about MySql ;)
[08:35] stagas has joined the channel
[08:35] EyePulp has joined the channel
[08:35] terinjokes: Been very successful in reinventing the wheel: http://image.bayimg.com/daombaacg.jpg
[08:35] felixge: (we are using it because we have relational data and our frontend is in CakePHP which favors mysql quite heavily)
[08:35] micheil: felixge: psst, aim small, then get big dude.
[08:35] zodiak has joined the channel
[08:35] sh1mmer: felixge: we are moving a lot of our stuff to a 'nosql' store internally
[08:36] sh1mmer: although there is a ton of mysql around still
[08:36] micheil: felixge: shoot me an email with a link to the source code and a brief quick intro: micheil@changelogshow.com
[08:36] _announcer: Twitter: "Coffeescript is actually looking nice, combine that with Node.js... *drool* http://bit.ly/boFlo7 http://bit.ly/cfl7pZ" -- Peter De Berdt. http://twitter.com/peterdeberdt/status/21299317987
[08:36] aubergine has joined the channel
[08:36] felixge: sh1mmer: I'm not saying there are no use cases for NoSQL, people have been abusing mysql as a key value store a lot in the past. But there is also relational data, and most of the NoSQL stores I've seen actually make me do more work for the same results as MySql
[08:36] sh1mmer: micheil: regarding step, I think it allows you to write readable code as you said, but I think you can write as easily legible code in an event-driven style
[08:37] micheil: true, but step can be / is event driven. although, I think it may be able to do it better.
[08:37] sh1mmer: micheil: part of the ways we'll make that successful is by defining ways to refactor sequential code that aren't evil.
[08:37] micheil: It's been a while since I looked at the source code.
[08:37] stride: felixge: any prerequesites on the db side for the test? I created a user with wildcard priviliges to an empty test database
[08:38] micheil: I'm more thinking of things where it's like: connect to database -> query database -> return result to user
[08:38] sh1mmer: micheil: I worry about abstractions like that though because they trick you into thinking sequentially
[08:38] felixge: stride: edit test/common.js
[08:38] felixge: stride: with your login credentials
[08:38] micheil: things that need to block because of third parties
[08:38] mattly has joined the channel
[08:38] stride: felixge: already done
[08:38] mape: felixge: how is the performance compared to other implementations? 
[08:38] decklin has joined the channel
[08:38] stride: Error: Expected call to connectCb() did not happen (process exit)
[08:38] micheil: sh1mmer: where each step depends on the previous step.
[08:38] felixge: mape: I have not benchmarked it, but the critical path for reading query results should be very fast
[08:38] stride: in test-client-connect.js
[08:38] felixge: mape: insert performance may not be as good as other drivers
[08:39] felixge: stride: seems like a timeout
[08:39] sh1mmer: micheil: right, but that's a fairly easy thing to code and I don't see any reason to not just use named functions
[08:39] felixge: stride: make sure your db is listening on port 3306
[08:39] micheil: eh, true.
[08:39] felixge: stride: localhost
[08:39] sh1mmer: micheil: I think we need some conventions for things like docs too
[08:39] micheil: sh1mmer: just saying, when you get into something like node-smtp-client, you can appreciate step.
[08:39] sh1mmer: micheil: people don't document their callbacks well enough
[08:39] micheil: 6 callbacks deep and all that.
[08:39] sh1mmer: heh
[08:40] micheil: (that was nov 09 code, thanks.)
[08:40] stride: felixge: don't have a mysql db on the same machine
[08:40] felixge: mape: but it should really be as fast as it can be for userland code, I spend a few thoughts on not wasting cycles :)
[08:40] felixge: stride: then you need to change the 'host' info
[08:40] sh1mmer: I've written code like that and I started to think about how I was doing all this anonymous stuff
[08:40] stride: felixge: ... :)
[08:40] jkreeftmeijer has joined the channel
[08:40] mAritz has joined the channel
[08:41] caolanm has joined the channel
[08:41] TrisMcC has joined the channel
[08:41] stride: I've put in the ip just in case it was a resolve-error, still the same results
[08:41] sYsk: what would be nice would be for node.js to spawn a new process just in case it detects the loop has been blocked for >Xseconds
[08:41] stride: mysql-client connects with the same parameters
[08:41] micheil: sh1mmer: it's more so because smtp is a sequential protocol.
[08:41] felixge: stride: what mysql version?
[08:41] pufuwozu has joined the channel
[08:42] sYsk: then you can see which requests cause node.js to block and refactor your code
[08:42] felixge: stride: I'm only 5.1.46 now, but I'm aiming for >= 4.1 compatibility
[08:42] sh1mmer: micheil: right, so I can see why step would be attractive, but I also think that it's easy to document the chain, or just follow it
[08:42] _announcer: Twitter: "Upgraded Node.js and can't for the life of me work out why I'm receiving "\x00\xC3\xBF" after sending "\x00\xFF". Odd!" -- Nathan de Vries. http://twitter.com/atnan/status/21299590978
[08:42] sh1mmer: if you don't use anonymous callbacks there isn't a big pyramid
[08:43] sh1mmer: and you can just write the callbacks source order according to execution order.
[08:43] Astro has joined the channel
[08:43] jspiros has joined the channel
[08:43] stride: felixge: client: "mysql  Ver 14.12 Distrib 5.0.77, for redhat-linux-gnu (x86_64) using readline 5.1" server: 5.0.45
[08:43] hoodow has joined the channel
[08:44] sh1mmer: micheil: I'm somewhat anti-abstraction when the abstraction is modelled off something different
[08:44] felixge: stride: that's the client, what about the server?
[08:44] felixge: stride: or is that what you get upon login?
[08:44] sh1mmer: micheil: we get this a lot with YQL where people don't understand why it doesn't work because they expect it to be SQL
[08:44] sh1mmer: and it can't be because it's making web service calls
[08:44] joshbuddy has joined the channel
[08:44] joshbuddy has joined the channel
[08:44] stride: felixge: "server: 5.0.45" last part of my message :)
[08:44] sh1mmer: but the convenience of the syntax lulls them into not understanding what they are doing.
[08:45] felixge: ah
[08:46] felixge: stride: is there any chance you could get me access credentials so I could debug?
[08:46] micheil: sh1mmer: there's also a point where nesting functions just gets too crazy.
[08:46] markwubben has joined the channel
[08:47] micheil: for instance, sure, write great code on the inside that nests callbacks as much as you like, but if I'm writing more then say 2-3 callbacks to send an email, I must be doing something wrong.
[08:47] micheil: (that's 2-3 callbacks after requiring the library)
[08:47] stride: felixge: nope, I abused a productive db for it, sorry. but I can setup a version on my nodejs test server, just a second
[08:47] _announcer: Twitter: ""How node.js saved my web application" http://bit.ly/925wiq &lt;-- opløftende." -- Ricco Førgaard. http://twitter.com/fiskeben/status/21299802845
[08:48] sh1mmer: micheil: yeah that sounds wrong
[08:48] teemow has joined the channel
[08:48] micheil: sh1mmer: that's a case where the public api should be simplified by making use of chaining to do common setup / related tasks.
[08:49] sh1mmer: micheil: right but chaining is also dangerous
[08:49] sh1mmer: it's one of the reasons code because hard to maintain
[08:49] felixge: stride: that'd be awesome
[08:51] micheil: uhh.. not chaining, I meant using something like step to link together those async methods to have just one callback, instead of say three
[08:51] sh1mmer: I think it depends on what you are doing
[08:54] beanie___ has joined the channel
[08:56] _announcer: Twitter: "I have a roll with node.js libev, libevent and add the option to choose their libev. However, libevent libev and I wonder which is better" [ja] -- Yuuki Takano. http://twitter.com/ytakanoster/status/21300178402
[08:56] micheil: sh1mmer: I'm thinking a end-developer api of this: http://gist.github.com/526659
[08:56] micheil: would be nice
[08:56] stride: felixge: strange, it's working locally, maybe I messed up the privileges or something
[08:57] felixge: stride: you should get a clean mysql error in that case
[08:57] rsms has joined the channel
[08:58] sh1mmer: micheil: that seems reasonable
[08:58] felixge: stride: actually ... I refactored the connect code, let me make sure it still handles errors properly
[08:58] micheil: sh1mmer: currently it's like, http://github.com/miksago/node-smtp-client/blob/master/examples/example-client.js
[08:59] stride: okay, just say the word and I'll test it again against the other db, I'll try to debug the serverside here to rule out errors on my side
[08:59] micheil: (imagine it as the new style callbacks though)
[08:59] sh1mmer: micheil: jeez.
[08:59] felixge: stride: no, there should be an exception thrown
[08:59] micheil: sh1mmer: really old code or what?
[09:00] felixge: stride: (the error object is not instanceof Error yet, but I'll fix that)
[09:00] sh1mmer: micheil: I think it's a stylistic thing
[09:00] Ori_P has joined the channel
[09:00] sh1mmer: there is a reason response.end() takes content
[09:00] tek has joined the channel
[09:00] sh1mmer: for http
[09:00] micheil: true
[09:00] micheil: but which would you prefer to use?
[09:01] sh1mmer: oh I agree
[09:01] sh1mmer: I think the former is much nicer
[09:01] micheil: so, the old way, or my new proposal?
[09:01] sh1mmer: I like objects of config personally
[09:02] sh1mmer: YUI uses that a lot
[09:02] stride: felixge: hm, yeah, just re-checked, not working on the remote db, gist of the full error output if you care: http://gist.github.com/526663
[09:02] micheil: the new style actually abstracts you from what smtp actually does somewhat, which is a good thing.
[09:02] sh1mmer: but at the same time having those raw events is still awesome
[09:02] micheil: sh1mmer: yeah, I know that. I've been using YUI to build Ideocase.
[09:02] micheil: just about to write my first YUI widget
[09:02] sh1mmer: :)
[09:03] micheil: sh1mmer: those raw events will still be there, trust me on that.
[09:03] felixge: stride: this would make sense if mysql is terminating the connection right away
[09:03] sh1mmer: right
[09:03] sh1mmer: but having a shorthand way to do stuff is important
[09:03] felixge: stride: or if there is a network error of some sorts
[09:03] stride: felixge: don't know, mysql-client works fine
[09:03] sh1mmer: I don't disagree
[09:03] felixge: stride: right
[09:03] micheil: sh1mmer: that's actually a project that I really need to give a lot of love to.
[09:03] rnewson has joined the channel
[09:03] rnewson has joined the channel
[09:03] felixge: stride: it's most likely to be a problem in my lib :)
[09:03] felixge: stride: but the only way to find out would be to test myself
[09:03] sh1mmer: micheil: I'm just adding the API to the DNS server so I'm doing this myself
[09:03] felixge: stride: or maybe you could do this:
[09:04] TomY has joined the channel
[09:04] felixge: stride: in client.js add this before line 56: p(packet);
[09:04] micheil: sh1mmer: I'm wishing I could go back and rewrite node-smtp-client, but time just doesn't permit at the moment, with having to do paid work and education at the same time.
[09:04] sh1mmer: life is like that
[09:04] felixge: stride: then run node test/system/test-client-connect.js
[09:04] sh1mmer: you aren't married
[09:05] stride: felixge: I can give you root access to the test db I set up, the problem is most likely reproducable when you connect to it
[09:05] sh1mmer: talk about a time sink :P
[09:05] micheil: sh1mmer: true. and I probably won't be for another 10 years.
[09:05] felixge: stride: sure, that would be even better :)
[09:05] sh1mmer: micheil: you never know, it sneaks up on you
[09:05] micheil: sh1mmer: context, i'm 17.75 now.
[09:05] stride: -> query in a second :)
[09:05] tav_ has joined the channel
[09:05] sh1mmer: I didn't expect to get married right after uni
[09:06] micheil: sh1mmer: you'd still be probably atleast 23-26 though
[09:06] sh1mmer: aye
[09:07] jblanche has joined the channel
[09:07] micheil: hmm.. time to write a quick bio.. horaay... not.
[09:07] tav__ has joined the channel
[09:08] sh1mmer: :)
[09:08] sh1mmer: I should sleep soon
[09:08] sh1mmer: I need to get up in um... 5 hours 20 minutes
[09:10] sveimac has joined the channel
[09:10] micheil: sh1mmer: be a real developer, it's not late until you need to get up in 1hr 10 minutes.
[09:11] micheil: (did that this morning >_> )
[09:11] sh1mmer: micheil: s/developer/hacker/
[09:11] micheil: :P
[09:11] sh1mmer: real developers go to work and get paid
[09:11] aubergine has joined the channel
[09:11] micheil: I joke, I joke.
[09:11] sh1mmer: by not missing their bus
[09:11] sh1mmer: micheil: I was young once
[09:11] micheil: still are, aren't you?
[09:12] sh1mmer: 29
[09:12] sh1mmer: I feel old
[09:12] micheil: seriously, when you're the age of my dad, you can say your old, until then, good luck :P
[09:12] sh1mmer: haha
[09:12] micheil: but seriously, in all truth, you are 11 years older than me.
[09:12] sh1mmer: it's young buggers like you being so bloody talented that makes me feel old
[09:13] guybrush: 
ACTION nods

[09:13] micheil: but really, we can joke about this stuff l;)
[09:13] micheil: sh1mmer: up by 6am? gee.. :P
[09:14] micheil: I only do that when I need to record a podcast with wynn or something like that.
[09:14] sh1mmer: 7.30 it's 2.15am now
[09:14] micheil: oho.. good point
[09:14] micheil: I thought of the 1 hr I said.
[09:18] mscdex: node.js rules!
[09:20] micheil: it doies.
[09:20] micheil: pity sleep isn't async, huh sh1mmer?
[09:20] sh1mmer: I keep meaning to go polyphasic
[09:22] micheil: lol.
[09:22] micheil: go the uberman sleep cycle, 6x20minute naps.
[09:22] pdelgallego: Is npm the standard node package manager (ala rubygems)? 
[09:22] micheil: you survive fine on only 2 hours of total sleep
[09:22] micheil: pdelgallego: defacto package manager
[09:24] sh1mmer: micheil: uberman is totally unsustainable
[09:24] sh1mmer: I was considering everyman though
[09:24] micheil: lol.
[09:24] micheil: everyman.. what's that one again?
[09:24] sh1mmer: it's brittle
[09:24] sh1mmer: core nap plus 3 short naps
[09:24] sh1mmer: well you can vary it
[09:26] pdelgallego: another question is there any emacs mode for nodejs ? 
[09:27] pdelgallego: I mean something to run tests,  run tasks, etc? 
[09:27] micheil: anyway chaps, I need to get to work, but if you release any awesome opensource stuff, shoot me an email: micheil@changelogshow.com
[09:29] omarkj has joined the channel
[09:31] aubergine has joined the channel
[09:34] digitalspaghetti: felixge: pint
[09:34] digitalspaghetti: ping
[09:38] _announcer: Twitter: "@atnan Don’t know anything about node.js, but 0xc3 0xbf is UTF-8 for 0xff UTF-16—i.e. it's a char. encoding issue." -- sutes. http://twitter.com/sutes/status/21302103656
[09:39] elliottkember has joined the channel
[09:39] felixge: digitalspaghetti: sup?
[09:39] digitalspaghetti: i can't get node-mysql to compile :(
[09:40] stride: compile? :)
[09:40] digitalspaghetti: stride: it has a makefile :)
[09:40] felixge: digitalspaghetti: that's just to run the tests
[09:40] felixge: digitalspaghetti: make test
[09:40] felixge: digitalspaghetti: no need to compile anything :)
[09:40] digitalspaghetti: ahh, so i can just stick it in .node_libraries ?
[09:41] felixge: digitalspaghetti: yeah, an NPM package is coming
[09:41] digitalspaghetti: cool
[09:41] digitalspaghetti: still getting used to node stuff :)
[09:41] felixge: digitalspaghetti: edit test/common.js with your db info to try it
[09:41] digitalspaghetti: gotcha
[09:42] micheil: felixge: wanna shoot me a quick email about this library?
[09:42] felixge: micheil: what do you need to know? The readme has all the docs, design goals, etc. :)
[09:42] micheil: s/library/module
[09:42] micheil: okay, just one with a link to it, so that I can follow-up in a few hours.
[09:43] micheil: micheil@changelogshow.com
[09:43] digitalspaghetti: i'm hoping this library solves this issues around why i didn't want to go with node, because i was being silly and picking a nosql backend :)
[09:43] digitalspaghetti: but i need a relational database
[09:43] Ori_P_ has joined the channel
[09:44] stride: dbslayer works fine for me in node
[09:44] _announcer: Twitter: "Blog update: Use node.js production http://is.gd/ejQPj nodejs # # # javascript daemon" [fr] -- Vincent RABAH. http://twitter.com/itwars/status/21302349035
[09:44] digitalspaghetti: doesn't dbslayer only work with selects though?  does it support inserts and updates now?
[09:44] stride: (just in case you don't want to wait)
[09:44] zomgbie has joined the channel
[09:45] stride: digitalspaghetti: don't know, sorry. I only use it for a monitoring component that only uses select
[09:46] felixge: micheil: emailed ya
[09:46] felixge: 
ACTION out for lunch

[09:46] stride: hm, dbslayer supports every sql you give it according to the docs, don't know about the node module
[09:48] mscdex: but dbslayer is a daemon and adds another dependency :/
[09:49] stride: yeah, true
[09:52] femtoo has joined the channel
[09:54] terinjokes: hrm... seems for every 1 load of my page, node is making 6 calls to the external site...
[09:59] _announcer: Twitter: "@jdub: JSON Template is available for node.js and Python too, in addition to in-browser-JS (and, of course it plays nicely w/ jQuery)" -- Garrett LeSage. http://twitter.com/garrett/status/21303022524
[09:59] mscdex: to what site?
[10:01] sveimac has joined the channel
[10:07] terinjokes: mscdex: i'm hitting graph.facebook.com for as a test
[10:08] terinjokes: adapting the example at http.Cleint so that its inside a http server
[10:11] _announcer: Twitter: "http://bit.ly/cUzTE1 SXSWi Panel proposition. #nodejs #nosql #nolimits #architecturesmarrantes http://yfrog.com/56q9wp" -- Ori A Pekelman. http://twitter.com/OriPekelman/status/21303609066
[10:12] stride: any chance you could share your code?
[10:13] terinjokes: ok course!
[10:14] terinjokes: http://pastebin.com/m1rKmBvj
[10:16] stride: oh, okay, you do that request to facebook everytime a request occurs, you should evaluate what exactly was requested from your server before doing that
[10:16] stride: e.g. every modern browsers checks for /favicon.ico and stuff
[10:16] terinjokes: stride: in production i would
[10:17] terinjokes: stride: i started node.js earlier this morning, and right now I'm just seeing how to do calls
[10:17] stride: well that accounts at least for one additional request
[10:17] terinjokes: one second...
[10:18] terinjokes: REQUEST: /favicon.ico
[10:18] stride: if (req.url == "/") { // foo }
[10:18] terinjokes: guess that solves that, haha
[10:19] mscdex: felixge: hrm... i do a 'make test' and i get an error: `test/{simple,system}/test-*.js': No such file or directory
[10:21] rnewson has joined the channel
[10:21] rnewson has joined the channel
[10:21] mscdex: felixge: also i don't have the 'gently' module :-(
[10:21] felixge: mscdex: npm install gently@latest
[10:22] mscdex: i don't have npm heh
[10:22] felixge: mscdex: http://github.com/felixge/node-gently
[10:22] _announcer: Twitter: "Being a graphic designer and programmer really pays off in moments like this - my C/Node.js tool has a name and lo http://twitpic.com/2fa268" -- Martin Rusev. http://twitter.com/martin_rusev/status/21304034497
[10:24] mscdex: felixge: seems to work fine now. cool!
[10:25] felixge: mscdex: awesome :)
[10:25] digitalspaghetti: felixge: so far so good on the mysql module :)
[10:26] felixge: digitalspaghetti: sweet :)
[10:26] felixge: I will probably start parsing dates into date objects and cast numbers / booleans properly
[10:26] felixge: unless you guys think getting the plain mysql 'strings' is best
[10:28] sveimac: thats a win
[10:30] mscdex: parsing dates would be nice... maybe have some way to toggle that functionality on/off somehow?
[10:30] mscdex: i dunno
[10:31] mscdex: not sure why i wouldn't want a date object
[10:31] mscdex: since you can easily tostring it
[10:32] MattJ has joined the channel
[10:32] mscdex: and yes, casting numbers and booleans properly would be great
[10:32] micheil: felixge: cool stuff
[10:34] sveimac: option with parse: false would be nice
[10:41] Ori_P has joined the channel
[10:44] shreekavi has joined the channel
[10:58] robotarmy has joined the channel
[11:01] felixge: mscdex: I'll probably add support for auto-casting into JS types which can be turned off
[11:04] _announcer: Twitter: "If you are interested in using #nodejs with #mysql, you may want to try out my new module: http://bit.ly/9cvg6g More details TBA soon" -- Felix Geisendörfer. http://twitter.com/felixge/status/21306054595
[11:04] danielzilla has joined the channel
[11:10] digitalspaghetti: felixge: i may have a bug, but unfortunatly i need to go out just now, won't be back for a couple of hours
[11:10] digitalspaghetti: what time will you be round till?
[11:10] felixge: digitalspaghetti: until 6pm, berlin time
[11:10] felixge: digitalspaghetti: feel free to report it as a github issue
[11:10] digitalspaghetti: yea, i'll be back before then
[11:10] digitalspaghetti: i dunno if it;s me being stupid though :D
[11:11] digitalspaghetti: (off to see David Mitchell at the Edinburgh Festival :)
[11:11] digitalspaghetti: anyway, better get off before i'm late :)
[11:11] exlibris has joined the channel
[11:12] daleharvey has joined the channel
[11:13] matt_c has joined the channel
[11:13] Gruni has joined the channel
[11:16] benburkert has joined the channel
[11:18] tuxsbro_ has joined the channel
[11:19] sideshowcoder has joined the channel
[11:21] feroz_ has joined the channel
[11:29] tmedema has joined the channel
[11:31] bsstoner has joined the channel
[11:34] tuxsbro has joined the channel
[11:41] sideshowcoder has joined the channel
[11:42] FransWillem has joined the channel
[11:46] phil- has joined the channel
[11:46] Tim_Smart has joined the channel
[11:47] jimmy203 has joined the channel
[11:47] Tim_Smart1 has joined the channel
[11:47] _announcer: Twitter: "I suspect a few bugs in out #gtugcampout game based on the #nodejs log. If your notice a bug: http://bit.ly/9o9MYd" -- Markus Knittig. http://twitter.com/mknittig/status/21308333809
[11:48] tmpvar has joined the channel
[11:48] jimmy203: I`ve used npm to install mongo - npm install mongodb bt I can`t seem to access mongod ? It says it is not installed? Any ideas?
[11:49] _announcer: Twitter: "I still dont like #javascript much, but #nodejs with websockets and #mongodb is so much #win." -- Almad. http://twitter.com/almadcz/status/21308430860
[11:54] Akufen has joined the channel
[11:57] fermion has joined the channel
[11:58] felixge has joined the channel
[11:58] felixge has joined the channel
[12:01] ben_alman has joined the channel
[12:05] _announcer: Twitter: "UglifyJS: a JavaScript parser/compressor/beautifier in #nodejs http://github.com/mishoo/UglifyJS" -- Aaron Heckmann. http://twitter.com/aaronheckmann/status/21309382529
[12:07] jetienne has joined the channel
[12:07] d0k has joined the channel
[12:08] sideshowcoder has joined the channel
[12:11] c4milo has joined the channel
[12:12] _announcer: Twitter: "Cutting-edge-web-dev land is all abuzz about #nodejs ... yet no one seems to have made the effort to write a set of passable docs for it." -- Marijn Haverbeke. http://twitter.com/marijnjh/status/21309840151
[12:14] gormer has joined the channel
[12:17] SubStack: Passable docs?
[12:17] SubStack: the api is pretty good last I checked
[12:17] SubStack: there's not very much to cover
[12:17] mscdex: the html docs need improvement
[12:20] Tim_Smart: docs are for noobs.
[12:20] Tim_Smart: >.>
[12:22] _announcer: Twitter: "Don't forget, @samshull is presenting Node.js tonight. NIIC @ 6:30. More details are on our site. Hope to see you there! http://j.mp/WHFU3" -- Screaming Monkeys. http://twitter.com/screaminmonkeys/status/21310418207
[12:23] SvenDowideit has joined the channel
[12:24] SubStack: just introspect the functionality out of everything
[12:24] stride: huh? what's wrong with the docs? formal explanation + examples for the base API and most modules
[12:25] SubStack: Object.getOwnPropertyNames(sys).map(function (name) { return sys[name].toString() })
[12:25] SubStack: there's your docs
[12:26] daleharvey: 
ACTION hasnt had a problem with node docs, I think they are pretty good overall

[12:27] daleharvey: I had to write new documentation for erlang because its sucked so much :P
[12:27] Tim_Smart: stride: I was just kidding. Docs are cool.
[12:30] gormer has joined the channel
[12:30] stride: Tim_Smart: I got that, I was referring to the tweet :)
[12:33] blogometer has joined the channel
[12:33] Tim_Smart: stride: Oh right. If someone needs more docs than what we already got, they probably shouldn't be developing.
[12:35] sveimac: Tim_Smart:  Thats abit harsh, theres alot of stuff that might need more examples
[12:35] Tim_Smart: A bit harsh - probably.
[12:35] daleharvey: " probably shouldn't be developing " + "with node.js"
[12:36] Tim_Smart: Unless they are new to programming as a whole I would be more forgiving :D
[12:36] Tim_Smart: What does need more docs, is the C++ v8 embedding API.
[12:37] omarkj: I can agree to that.
[12:37] aheckmann has joined the channel
[12:37] mscdex: stride: there's lots of things that could be improved, such as showing "inheritance", showing better/more examples for the more complex things, etc
[12:38] sveimac: I think the #node.js community should asume that there are alot of frontend people who arent used to backend stuff 
[12:39] femtoo has joined the channel
[12:40] mscdex: stride: also, generating better and more predictable formats for links to sections/subsections is needed
[12:40] daleharvey: sveimac: and I think people who arent particularly experienced developers should assume that node if a new project thats under heavy dev and not particularly newbie friendly, yet
[12:40] daleharvey: *node is a new project
[12:40] voodootikigod has joined the channel
[12:41] phiggins has joined the channel
[12:41] sveimac: sure agree, thats a general for something thats not even close to 1.0 :)
[12:41] kevm_ has joined the channel
[12:41] omarkj: I think we should just be friendly to all questions, isn't that easiest ?
[12:41] ben_alman has joined the channel
[12:42] mscdex: omarkj: yes, but there are lots of questions and things that could have been easily remedied by better html documentation imho
[12:42] sveimac: just not lets get to smart about it :)
[12:42] omarkj: mscdex: True.
[12:42] mscdex: lots of common questions anyway
[12:43] phiggins: which one of you is the 4chan fan http://twitter.com/voodootikigod/status/21282969378
[12:43] mscdex: never heard of it
[12:44] _announcer: Twitter: "why i like node.js over eventmachine: node.js actually works when you write something only theoretically, em doesn't. also em is more opaque" -- Sam Elliott. http://twitter.com/Lenary/status/21311831975
[12:45] stride: mscdex: all your points are true but it is certainly sufficient to get started or dive into for more complex issues
[12:46] omarkj: That being said, I wasn't aware of the community being very hostile.
[12:46] daleharvey: someone mentioned the word "ruby fag" in passing and that whole 4chan thing went off, was a tad overreactive
[12:47] phiggins: i just saw the tweet is all
[12:48] phiggins: also node-xml parser is retarded.
[12:48] evlapix has joined the channel
[12:48] omarkj: Ah.
[12:48] mikeal has joined the channel
[12:48] jherdman has joined the channel
[12:48] phiggins: it dies on <script>var a = "<div></div>";</script> fatal-don't-keep-parsing-style.
[12:49] stride: so?
[12:49] daleharvey: why would it parse that if its an xml parser?
[12:49] voodootikigod_ has joined the channel
[12:49] blowery: QUICK EVERYONE JUMP ON PHIGGINS
[12:49] daleharvey: lol
[12:49] phiggins: i'm looking for a general purpose dom parsing lib
[12:49] phiggins: blowery is in fact a ruby fag
[12:50] blowery: only the best for me
[12:50] blowery: gem flavored cigarettes.
[12:50] daleharvey: I havent used any to suggest, theres a few in http://wiki.github.com/ry/node/modules though
[12:51] voodootikigod_: phiggins: apricot
[12:51] voodootikigod_: phiggins: http://github.com/silentrob/Apricot
[12:51] kriszyp has joined the channel
[12:51] phiggins: i could just hack it to treat <script> as cdata I guess
[12:52] phiggins: hmm
[12:52] phiggins: apricot looks promising.
[12:52] paul___ has joined the channel
[12:52] phiggins: god do I have to use sizzle. 
[12:52] phiggins: ;)
[12:52] voodootikigod: rob is one of the nitobi guys
[12:53] voodootikigod: err
[12:53] voodootikigod: was
[12:53] voodootikigod: now with joyent
[12:53] voodootikigod: pretty sure you have met him
[12:53] mscdex: phiggins: that's not valid xml
[12:53] blowery: phiggins: god loves sizzle
[12:53] mscdex: phiggins: the < and > have to be escaped
[12:53] mscdex: er encoded rather
[12:54] phiggins: mscdex: i know but 99.99999% of the interweb has something like that going on
[12:54] phiggins: and I don't need it to be valid xml, i need it to parse html
[12:55] mscdex: well, html is a different beast :p
[12:55] phiggins: not by much really
[12:55] phiggins: i'm not worried about the validation, just the ability to traverse and grab attributes and whatnot
[12:55] mscdex: for scraping i generally run the html through tidy, converting it to xml in the process, and then using the likes of libxmljs to get at the data
[12:56] phiggins: shopstopper for me if it dies on var template = "<div ${stuff}></div>"
[12:56] mscdex: do you have a url for the page you're trying to get at? also, are you using this for scraping or something else?
[12:57] phiggins: processing local files
[12:57] mscdex: oh
[12:57] phiggins: php's DOMDocument _works_ but it mangles lots of the markup
[12:57] phiggins: and forces case sensitivity on attributes
[12:57] phiggins: er insensitivity
[12:58] davidc_: phiggins, what are you doing with DOMDocument exactly?
[12:58] phiggins: anyway it lowercases all attributes in it's parsing, and there is a disconnect between the spec and way the world works.
[12:58] davidc_: You might be better of simply with XMLWriter (Faster, more flexible, better API)
[12:58] phiggins: i've given up on the notion of being able to write this back out
[12:59] phiggins: i'm using it to report things. i guess I could give a hair of background:
[12:59] voodootikigod_ has joined the channel
[12:59] phiggins: i deprecated the dojoType="foo.Bar" attribute for the dojo parser
[12:59] mscdex: phiggins: there's also jsdom and a node html parser
[12:59] phiggins: and want to help make a migration thing for folks
[12:59] phiggins: mscdex: yah voodootikigod's link includes those two with a wrapper it looks like
[13:00] voodootikigod: voodootikigod is only mildly knowledgeable
[13:00] margle has joined the channel
[13:00] mscdex: huhu
[13:01] phiggins: i can regexp replace dojoType="bar" to data-dojo-type="bar" but the issue is that "all the other nonstandard props are shoved in a single json string" on data-dojo-props
[13:01] phiggins: so I need node[dojoType] and then build up json of all the other attributes
[13:02] phiggins: anyway just found that I couldn't find a capable html-parser over the weekend
[13:03] phiggins: and php's $doc -> loadhTMLFile($file); $doc -> saveHTMLFile($file); makes a really messy diff
[13:04] mscdex has joined the channel
[13:06] xiay has joined the channel
[13:07] nerdEd has joined the channel
[13:07] wattz: Good morning
[13:07] mscdex has joined the channel
[13:08] kodisha has joined the channel
[13:11] _announcer: Twitter: "lisp.js: a Lisp for node.js - lisp-js's posterous http://bit.ly/acS3p8" -- Web Developer Links. http://twitter.com/webdevlinks/status/21313571070
[13:15] astrolin has joined the channel
[13:16] mischief has joined the channel
[13:17] daleharvey has joined the channel
[13:17] geeks_bot1 has joined the channel
[13:18] zapnap has joined the channel
[13:20] astrolin has joined the channel
[13:21] _announcer: Twitter: "reading Node.js code reminds me of an fundamentalist  Inversion of Control code." -- Josh Robb. http://twitter.com/josh_robb/status/21314291631
[13:21] _announcer: Twitter: "Playing with node.js has shown that I didn't know as much about HTTP as I thought I did." -- Ricky Hussmann. http://twitter.com/rhussmann/status/21314293259
[13:23] maushu has joined the channel
[13:23] maushu: Oh, f*ck me sideways.
[13:24] mbleigh has joined the channel
[13:25] stride: I'd rather not
[13:26] shreekavi has left the channel
[13:27] maushu: godaddy is like the source of all evil.
[13:27] alexkwolfe has joined the channel
[13:28] mape: But they have the cheapest domains!
[13:28] codemariner has joined the channel
[13:28] mape: And models strutting around
[13:29] mape: That has to count for something
[13:29] skampler: and hot chicks on the frontpage, that's a deal breaker for me
[13:29] jtsnow has joined the channel
[13:31] sideshowcoder has joined the channel
[13:35] gormer has joined the channel
[13:38] charlesjolley- has joined the channel
[13:39] omarkj: If they have hot chicks on the frontpage, I'll buy from them.
[13:40] SamuraiJack_ has joined the channel
[13:40] dylang has joined the channel
[13:41] davidwalsh has joined the channel
[13:42] gormer has joined the channel
[13:42] phiggins has joined the channel
[13:43] pgriess has joined the channel
[13:44] cce: maushu: what did they do?
[13:46] maushu: cce: Besides upselling and having the WORST admin panel ever?
[13:46] cce: ah, yea, hidden-cost-of-crap
[13:46] cce: 
ACTION winks.

[13:47] steadicat has joined the channel
[13:49] Akufen has joined the channel
[13:49] Akufen has joined the channel
[13:54] aubergine has joined the channel
[13:55] fearphage has joined the channel
[13:56] xla has joined the channel
[13:58] nerdEd has joined the channel
[13:58] mscdex: maushu.com isn't taken! :O
[13:58] mikeal has joined the channel
[13:58] maushu: mscdex: woot.
[13:59] maushu: Go ahead, make my day punk.
[13:59] maushu: I will sue you to death.
[14:00] maushu: Besides, who the hell gets domains based on their irc nickname?
[14:00] stride: you should file a patent on nickname based domain names first
[14:01] jakehow has joined the channel
[14:02] mscdex: maushu: lots of people on here do
[14:07] daleharvey: lol
[14:08] daleharvey: The main idea of "Inception": if you run a VM inside a VM inside a VM inside a VM, everything will be very slow
[14:08] daleharvey: do, wrong room, I am blaming that on maushu 
[14:08] mscdex: omghax
[14:08] piotrekm has joined the channel
[14:10] piotrekm: Hi, is there anyone capable of editing the documentation? http://nodejs.org/api.html#stream-writeable-37 - I think it should read 'writable', as this is the property name here for example: http://pastebin.com/v5ezMDqx
[14:12] daniellindsley has joined the channel
[14:12] softdrink has joined the channel
[14:13] fearphage has joined the channel
[14:13] bryanl has left the channel
[14:13] ben_alman has joined the channel
[14:14] matt_c has joined the channel
[14:14] mscdex: piotrekm: you can post on the mailing list about it and it should get accepted
[14:14] mscdex: especially if you supply a patch
[14:15] piotrekm: ok
[14:16] dipser has joined the channel
[14:16] maushu: piotrekm: clone the repository, change docs, create patch, send patch.
[14:16] astrolin has joined the channel
[14:16] piotrekm: thanks
[14:17] maushu: The almight node.js gods will check the worthiness of your patch and strike it down if its not worthy.
[14:20] wllm has joined the channel
[14:22] alexkwolfe has joined the channel
[14:24] _announcer: Twitter: "Cygwin, CoffeeScript & Growl for Windows integration - Let’s start by adjusting where node.js looks for... http://tumblr.com/xskfx8sqn" -- Stan Angeloff. http://twitter.com/StanAngeloff/status/21318760852
[14:24] ThePub has joined the channel
[14:25] stagas_ has joined the channel
[14:26] mscdex: maushu: that's not very encouraging
[14:26] _announcer: Twitter: "@follesoe there also is Haml.js I believe use with Node.js and also a MongoDB.js mapper" -- Mark Nijhof. http://twitter.com/MarkNijhof/status/21318921643
[14:27] JimBastard has joined the channel
[14:27] [[zz]] has joined the channel
[14:29] piotrekm: and, still brooding upon writability, does stream.end(data) also need it checked, or does it check writable state itself?
[14:33] aliem has joined the channel
[14:33] mscdex: piotrekm: http://github.com/ry/node/blob/master/lib/net.js#L1024
[14:35] piotrekm: mscdex: thanks again
[14:35] mscdex: np
[14:41] behmann has joined the channel
[14:46] matt_c: phiggins: I'm sure you've already found this out, but Apricot is pretty sweet.
[14:47] phiggins: jep voodootikigod pointed me in that direction, will take a look after tehdayjob
[14:47] rwaldron has joined the channel
[14:48] confoocious has joined the channel
[14:48] _announcer: Twitter: "So, Do do do signature generation in the state of our Bash is unreasonable. At such time for the node.js. The token made from node.js bash curl up with a good feeling it 飛Base. In dealing with the troublesome post node.js" [ja] -- こば@変人生徒会長. http://twitter.com/KOBA789/status/21320544491
[14:49] aubergine has joined the channel
[14:50] matt_c: Parsing RSS with CSS selectors is more fun than it sounds.
[14:52] drudge: anyone using node-inspector?
[14:52] christophsturm has joined the channel
[14:53] softdrink: beats xpath
[14:55] loincloth has joined the channel
[14:55] ph^ has joined the channel
[14:55] JimBastard has joined the channel
[14:56] shockie has joined the channel
[14:56] ajpiano has joined the channel
[14:56] jherdman has joined the channel
[14:58] dnm has joined the channel
[14:59] hdon has joined the channel
[15:00] fod has joined the channel
[15:03] SingAlong has joined the channel
[15:04] SingAlong: hi all
[15:04] SingAlong: how can I access the localserver on my computer from another computer (connected via the same router)? I'm using node.js and it seems i'll have to bind the address/port when i create the server
[15:04] atmos: set the ip to 0.0.0.0 and you should be able to hit the internal ip on your network, 192.168.x.x
[15:10] brapse has joined the channel
[15:10] sonnym has joined the channel
[15:11] jetienne has joined the channel
[15:12] vdrab has joined the channel
[15:17] cardona507 has joined the channel
[15:17] feydr has joined the channel
[15:17] cardona507: w00t
[15:19] ryankshaw has joined the channel
[15:19] margle has joined the channel
[15:19] ryankshaw has joined the channel
[15:21] _announcer: Twitter: "#NodeJS anyone know how it stands up to other server side languages??" -- Jason Gorman. http://twitter.com/jsongorman/status/21322891670
[15:21] bradleymeck has joined the channel
[15:24] SingAlong: atmos: thanks! that worked.
[15:25] SingAlong: which the best node pkg to serve static content? I found a bunch of them including node-paperboy and node-static
[15:25] robinduckett has joined the channel
[15:25] stagas__ has joined the channel
[15:26] _announcer: Twitter: "I didn't know about node.js then @iancmyers and @raycmorgan told me about it. Then node-mysql alpha was announced. I'm about to do something" -- Dave Stevens. http://twitter.com/shakefon/status/21323193654
[15:26] rsms has joined the channel
[15:27] FransWillem: Did anyone ever attempt to write an NNTP protocol implementation in Node.js ?
[15:27] vdrab: does the node file system api currently support random access to files, like seek()'s ?
[15:27] [[zz]] has joined the channel
[15:28] ehaas has joined the channel
[15:30] TrueShot has joined the channel
[15:34] FransWillem: vdrab: yes
[15:34] daniellindsley has joined the channel
[15:35] tisba has joined the channel
[15:35] softdrink: has anyone adopted sass's nested output format as their 'standard' way of doing normal css? just curious....
[15:38] jbrantly has joined the channel
[15:40] tmpvar has joined the channel
[15:40] _announcer: Twitter: "http://bit.ly/aIHLc5 ¦ Mysql for NodeJs ! #node #mysql" -- François-G. Ribreau. http://twitter.com/FGRibreau/status/21324217971
[15:40] mjr_: FransWillem: I think mikeal has NNTP in node.
[15:40] _announcer: Twitter: "lisp.js, cause coffeescript is for pussies :P http://lisp-js.posterous.com/lispjs-a-lisp-for-nodejs-0" -- Mikeal. http://twitter.com/mikeal/status/21324237240
[15:40] mikeal: i have part of an implementation
[15:41] devinus: lol @ lisp.js
[15:41] matt_c: 
ACTION has done parts of gopher and finger in node.  It's pretty fun for network protocols.

[15:41] mikeal: basically, it's enough for me to script some stuff than pull down the bin parts of alt.bin.comics.dcp and stick them in CouchDB :)
[15:42] mikeal: NNTP is a pain in the ass
[15:42] mikeal: it's insanely stateful
[15:42] mjr_: matt_c: how is your WAIS implementation coming?
[15:42] mikeal: and that's annoying to parse in node
[15:44] FransWillem: mikeal: Is yours open source somewhere? I'm currently writing my own implementation (basically want to index some newsgroups and automatically pull some NZBs from it), would love to see how you did it
[15:45] _announcer: Twitter: ""Nitrode: A lightweight HTTP Server on top of Node.js" by @changelogshow #tech #javascript http://bit.ly/aF4zdq" -- Elijah Manor. http://twitter.com/elijahmanor/status/21324605312
[15:46] dnm has joined the channel
[15:47] shockie has joined the channel
[15:50] stephank has joined the channel
[15:50] drudge: http://github.com/mikeal/node-nntp
[15:51] sh1mmer has joined the channel
[15:51] FransWillem: Eeeh, that's horrible :p
[15:52] sveimac: softdrink: I have some, neat with big projects at least 
[15:53] stepheneb has joined the channel
[15:55] kodisha has joined the channel
[15:55] mu-hannibal has joined the channel
[15:55] drudge: FransWillem: you scared for life? :P
[15:55] drudge: scarred* even
[15:55] stephank: Hello! I'm trying to write a buffer to a file with writeFileSync, but it throws. This is what I'm trying: require('fs').writeFileSync('test.out', new Buffer([1,2,3]));
[15:55] stephank: Is there something I'm doing wrong there?
[15:56] nrstott has joined the channel
[15:56] nrstott has joined the channel
[15:56] FransWillem: what does it throw?
[15:56] stephank: TypeError: Bad argument.  at Object.writeFileSync (fs:400:17)
[15:58] atmos: prolly shouldn't be a buffer there
[15:58] atmos: what if you give it a string /
[15:58] stephank: well, that works, but I want to write binary data
[15:58] caolanm: hmm, pretty sure it should take a buffer
[15:58] caolanm: but it throws for me too
[15:59] exlibris has joined the channel
[15:59] caolanm: its calling slice on the buffer I think
[16:00] caolanm: which breaks
[16:00] bradleymeck: yea, buffer doesnt do a slice op that is just generic :(
[16:01] caolanm: but writeSync looks like it will take a buffer
[16:01] caolanm: so perhaps writeFileSync is a bit out of date?
[16:02] sh1mmer: stephank: I don't think that argument can be an array
[16:02] femtoo has joined the channel
[16:02] sh1mmer: sorry, I don't think the argument for the buffer can be an array
[16:02] atmos: yeah, the docs appear to be wrong
[16:03] stephank: sh1mmer: the buffer can definitely be initialized with an array.
[16:03] atmos: sh1mmer: try it in node-repl, it works fine
[16:03] stephank: I'm going to try with writeSync.
[16:03] sh1mmer: ok
[16:03] sh1mmer: I was trying now
[16:03] sh1mmer: hm, that's cool I didn't know you could do that
[16:04] sveisvei has joined the channel
[16:04] stephank: Looks like it's working with writeSync, thanks. :)
[16:05] caolanm: interesting
[16:05] Aria has joined the channel
[16:05] stephank: Should I open an issue for writeFileSync?
[16:06] caolanm: Best document it somewhere, so it can be looked at by someone with a better understanding than me ;)
[16:06] sh1mmer: seems like
[16:06] caolanm: GitHub issues or post on the mailing list?
[16:06] stephank: I'll create an issue in github.
[16:08] calebc has joined the channel
[16:10] calebc: I have a question about Express. Is there an IRC channel for it?  If not, is it appropriate to ask it here?
[16:10] _announcer: Twitter: "Ran out of time to put it node.js server, so for now try to imac without sampling day was started." [ja] -- Yohei Sasaki. http://twitter.com/yssk22/status/21326419035
[16:11] wllm has joined the channel
[16:11] jakehow has joined the channel
[16:12] charlesjolley- has joined the channel
[16:13] deepthawtz has joined the channel
[16:15] _bastardbot has joined the channel
[16:16] JimBastard has joined the channel
[16:16] JimBastard: !tweet test
[16:16] piotrekm: Is there a chance 'data' events will overlap, so the assigned function will be called second time while the first call is still running?
[16:17] digitalspaghetti: felixge: you still around?
[16:17] JimBastard: !tweet test
[16:17] JimBastard: aight hes online
[16:17] felixge: digitalspaghetti: y
[16:17] felixge: digitalspaghetti: but about to head out for beach volleyball
[16:17] claudiu__ has joined the channel
[16:17] digitalspaghetti: ahh, k - i might just post this to the mailing list then
[16:19] mjr_ has joined the channel
[16:21] caolanm: piotrekm: no, unless the callback does some other async operation which may still be in progress when the second data event is handled
[16:21] mikeal has joined the channel
[16:21] bpot has joined the channel
[16:22] FransWillem: ...
[16:22] FransWillem: How come I always end up implementing more than I actually need ?
[16:22] caolanm: piotrekm: a callback will not be called twice at the same time because node is single threaded
[16:22] FransWillem: Somehow I just need login and overview commands for my NNTP implementation, yet ended up with Head,Body,Article,Stat,Last,Next,Date,etc
[16:23] AndIrc_ has joined the channel
[16:23] SingAlong: how can I install this node pkg called node-paperboy. (to serve static files)
[16:23] kodisha_ has joined the channel
[16:23] SingAlong: doing npm install node-paperboy doesnt find it
[16:23] piotrekm: caolanm: so e.g. if function called on data event is hanging for some time, even the 'timeout' event won't be triggered?
[16:23] stephank has left the channel
[16:24] caolanm: piotrekm: the timeout might be triggered before or after the next data event, but you will not have 2 pieces of code running at the same time
[16:24] piotrekm: caolanm: thanks
[16:25] robotarmy has joined the channel
[16:25] mjr_: SingAlong: I think the package in npm is just called 'paperboy'
[16:26] mjr_: SingAlong: npm ls paper
[16:26] SingAlong: mjr_: does the express framework serve static files? if it does then i dont need paperboy
[16:26] robotarmy has joined the channel
[16:26] caolanm: SingAlong: you might be able to use Connect to server static files
[16:26] caolanm: I do that a lot
[16:26] mjr_: I've never actually used it, but I believe it does.
[16:26] caolanm: and express is built on connect
[16:26] SingAlong: caolanm: I need to use express too :)
[16:27] caolanm: you can use connect middleware with express I think
[16:27] caolanm: but I'd be surprised if express had nothing for serving static files
[16:28] mjr_: I keep hearing people talk about a "static provider", so that's probably it.
[16:28] mjr_: One of these days, I'll do something with Connect.  Sounds really cool.
[16:30] mikeal has joined the channel
[16:30] dmcquay has joined the channel
[16:30] stride: mjr_: yeah, that's it
[16:30] jimt has joined the channel
[16:31] joshholt has joined the channel
[16:32] jchris has joined the channel
[16:32] jchris has joined the channel
[16:35] joshbuddy has joined the channel
[16:36] SingAlong_ has joined the channel
[16:41] jchris has joined the channel
[16:43] c4milo has joined the channel
[16:45] ajpiano has joined the channel
[16:47] SingAlong: node.js says sendfile isnt defined
[16:47] SingAlong: (object has no methd sendfile)
[16:48] _announcer: Twitter: "wow ... use event loop instead of threads to deal with IO for server scalability ... that makes total sense! http://nodejs.org" -- Peter Chang. http://twitter.com/peterchang/status/21329053320
[16:49] mscdex: SingAlong: what version of node are you using?
[16:49] mu-hannibal has joined the channel
[16:50] mscdex: SingAlong: http://github.com/ry/node/blob/master/lib/fs.js#L276
[16:51] SingAlong: mscdex: 0.1.99-39-gd49d53f
[16:51] maushu: Maybe you are thinking of the stream.write(string, encoding='utf8', [fd])
[16:51] mscdex: SingAlong: try updating node... i'm not sure when sendfile was added
[16:52] maushu: Or that.
[16:52] EyePulp has joined the channel
[16:52] zodiak has joined the channel
[16:52] maushu: ryah is the biggest troll ever. The api can change as jell-o in a earthquake.
[16:53] maushu: We are going to reach 0.2 in 2020.
[16:53] mscdex: hrm... i see fs.sendfile in node on april 8
[16:53] mscdex: at least
[16:54] tango3 has joined the channel
[16:54] mr_danie1: I don't know if the question makes sense: is there a way to print/show the current JavaScript event stack in a browser or a Node.js environment?
[16:54] mr_danie1: if I understand correctly then JavaScript is event based, executing one event after the other. And for this some sort of queue is needed, which executed the events in a specific order
[16:55] mscdex: mr_danie1: in node: (new Error()).stack
[16:55] fod has joined the channel
[16:55] WALoeIII has joined the channel
[16:55] mscdex: mr_danie1: as far as the browser goes, i know mozilla has some proprietary stuff that you can get at
[16:56] mscdex: not sure about what to use for other browsers though
[16:56] charlenopires has joined the channel
[16:57] SingAlong: http://pastie.org/1095843 anything wrong there?
[16:57] SingAlong: I'm using express
[16:57] deepthawtz: npm tells me i should bug the author of mjsunit to publish it, so it that is you, publish it.
[16:57] mscdex: mr_danie1: for browsers, try this maybe: http://eriwen.com/javascript/js-stack-trace/
[16:57] mr_danie1: browser is not important: Firefox, Chrome, ... whatever. I only need a way to get the current event stack to understand JavaScript better
[16:57] _announcer: Twitter: "Working through an "old" node.js and express tutorial can be painful, but I'm learning a lot by fixing it!" -- Caleb Cornman. http://twitter.com/calebcornman/status/21329678728
[16:57] mscdex: mr_danie1: i'd try that link or use (new Error()).stack from node
[16:57] maushu: Hmm, to get the stack using a debugger would be a good idea.
[16:58] maushu: You can then go step by step and see the stack change.
[16:58] SingAlong: anyone on http://pastie.org/1095843 ?
[16:58] stride: SingAlong: what's that weird "ƒ" character?
[16:59] mscdex: SingAlong: i don't know what the this context is there... but sendfile is a method of the fs module
[16:59] mscdex: so: fs.sendfile(...);
[16:59] SingAlong: stride: emacs shortens "function" dont mind that
[16:59] maushu: mscdex: That doesn't seem raw node.js
[16:59] stride: ah, ok
[16:59] mscdex: maushu: huh?
[16:59] jetienne_ has joined the channel
[17:00] maushu: The this is a reference to the request and the sendfile is probably a function to send the file to the requester.
[17:00] maushu: SingAlong: Try "sendFile".
[17:00] voxpelli has joined the channel
[17:00] maushu: bbl
[17:00] mscdex: no idea
[17:00] mscdex: that's kind of confusing heh
[17:00] maushu: 
ACTION facepalms.

[17:01] SingAlong: maushu: I'll try that in a min. updating node right now.
[17:01] mscdex: 
ACTION palmsface

[17:01] maushu: 
ACTION facepalms' mscdex.

[17:01] mscdex: ;()
[17:01] rauchg_ has joined the channel
[17:01] twoism has joined the channel
[17:02] samsonjs has joined the channel
[17:03] samsonjs has joined the channel
[17:04] sh1mmer has joined the channel
[17:04] samsonjs has joined the channel
[17:05] samsonjs has joined the channel
[17:07] samsonjs has joined the channel
[17:07] amerine has joined the channel
[17:08] amerine has joined the channel
[17:09] programble has joined the channel
[17:10] jblanche has joined the channel
[17:10] FransWillem has joined the channel
[17:11] _announcer: Twitter: "A lightweight, super fast HTTP server built on Node.JS http://post.ly/s7iD" -- Ivan Ribeiro. http://twitter.com/irr/status/21330669052
[17:17] JimBastard: these all day meetings are fucking brutal
[17:17] JimBastard: if i dont write some code soon im gonna lose it
[17:18] astrolin_ has joined the channel
[17:21] derferman has joined the channel
[17:21] dispalt has joined the channel
[17:21] benburkert has joined the channel
[17:22] _announcer: Twitter: "Another nice project with PEG.js-based parser: lisp.js. http://lisp-js.posterous.com/lispjs-a-lisp-for-nodejs-0" -- PEG in JavaScript. http://twitter.com/peg_js/status/21331388353
[17:26] _announcer: Twitter: "In the lecture on the CEP # TDC2010 summer of toast a little Node.JS will show a CEP engine with Javascript" [pt] -- lucabastos. http://twitter.com/lucabastos/status/21331642954
[17:26] Throlkim has joined the channel
[17:27] aubergine has joined the channel
[17:29] mr_danie1: mscdex: it seems like that 'new Error().stack' only prints the current stack, but not the current state of the 'event queue' 
[17:29] sudoer has joined the channel
[17:29] sh1mmer: what do people think about the idea of adding some notion of bubbling and propigation to eventemitter ? 
[17:30] JimBastard: sh1mmer: naaah
[17:30] JimBastard: not in core
[17:30] JimBastard: imho
[17:31] rauchg_ has joined the channel
[17:32] blogometer has joined the channel
[17:33] mikeal has joined the channel
[17:33] stagas has joined the channel
[17:33] blogometer: Someone on Twitter says ya'll are like 4chat.
[17:33] blogometer: 4chan.
[17:34] grahamalot has joined the channel
[17:34] blogometer: But, you're all hugs and kisses to me. What's up with that?
[17:34] davidc_: haha :)
[17:34] blogometer: Are you being mean when I'm not looking?
[17:34] jackish has joined the channel
[17:34] mscdex: mr_danie1: you can't get at what's currently on the event loop from js
[17:35] creationix has joined the channel
[17:38] rnewson: so this blows up (http://friendpaste.com/4urOunGauc1GOV7jXJ46SG) if the server returns an error status, but doesn't if it succeeds. I think it's a node.js bug. the request is a put and unconditionally writes a request body. If I don't pass a body, it works fine.
[17:39] aheckmann has joined the channel
[17:40] inkubus08 has joined the channel
[17:41] jamescarr_: damn
[17:42] sh1mmer: rnewson: define 'blows up
[17:42] rnewson: the node.js process exits with Broken pipe.
[17:42] sh1mmer: not an exception?
[17:43] rnewson: I'll show you as soon as friendpaste wakes up
[17:43] rnewson: http://friendpaste.com/4QN5AS6X7o1Xn0iutUWfJ3
[17:43] sh1mmer: heh
[17:43] sh1mmer: use gists
[17:43] sh1mmer: nom
[17:44] sh1mmer: that seems like a bug, that should really be an exception
[17:44] skampler: rnewson: node version?
[17:44] rnewson: latest, built yesterday.
[17:45] rnewson: wait, on this box it's 0.1.104 from homebrew
[17:45] rnewson: the lab has one from source which is about as new
[17:45] ink83 has joined the channel
[17:45] rnewson: at a guess, the connection is closed but node.js writes the body anyway.
[17:45] rnewson: I'm not sure what the http rules are here, actually. gets a bit sketchy
[17:46] mscdex: rnewson: i'm not sure you can reuse a client like that, can you?
[17:46] rnewson: the whole expect/continue thing exists to avoid sending the body when the server is going to return an error.
[17:46] rnewson: so my guess is that without it, the server is required to swallow it
[17:46] rnewson: mscdex: when nodeload does it it uses one client, I can verify that right now
[17:46] mscdex: ok
[17:47] rnewson: or do you mean *not* reuse the client object?
[17:47] rnewson: now it gets the other error: 'ECONNREFUSED, Could not contact DNS servers' because it's massively parallel. :)
[17:48] sh1mmer: rnewson: 8.2.4 http://www.w3.org/Protocols/rfc2616/rfc2616-sec8.html
[17:48] rnewson: running out of fd's in this case, but 100,000 connections is silly.
[17:48] sh1mmer: "If at any point an error status is received, the client
[17:48] sh1mmer:       - SHOULD NOT continue and"
[17:49] sh1mmer: "- SHOULD close the connection if it has not completed sending the
[17:49] sh1mmer:         request message."
[17:49] rnewson: ah, ok. so node.js closes the connection but still tries to write, which is what it looks like.
[17:49] sh1mmer: without looking at the source that would be my guess too
[17:49] mscdex: rnewson: if you use .end(new Buffer()) to write, you won't have to worry about checking the writable property
[17:50] rnewson: mscdex: the writable property? I'm not familiar. is it new?
[17:51] mscdex: no
[17:51] rnewson: I'm supposed to check it before I write?
[17:52] rnewson: this is an extracted test to see if it's a node.js bug. the real code is inside nodeloadlib.js, an http perf. framework I'm using.
[17:54] sh1mmer: rnewson: I think in general the http client needs an exception for this beyond just the exception from the write stream
[17:54] sh1mmer: because it's a complexity of HTTP
[17:54] rnewson: yes.
[17:55] rnewson: but the connection should be closed and replaced without blowing up in the client code, yes? 
[17:57] rnewson: that is, my write(buf) shouldn't fail. I should see that it didn't work when my resp.statusCode is 409 or whatever.
[17:57] tekky has joined the channel
[17:58] sh1mmer: I think it should be caught and a new exception should be thrown
[17:58] sh1mmer: which you should handle
[17:58] mscdex: rnewson: no, you only need to check the property if you don't use .end() to write your data
[17:59] rnewson: testing if (req.writeable) really helps, thoug.
[17:59] mscdex: .end() automatically checks .writeable
[17:59] mscdex: is the reason
[17:59] rnewson: mscdex: ok, cool. calling end(date) wouldn't  work for me as nodeload calls end for me (it adds handlers to trap the response for its internal monitory).
[17:59] mscdex: ah i see
[18:00] _announcer: Twitter: "@ L0utus TwitterBot artificial system is incapable of making me a target, BASIC, so I cut to the implementation of the OAuth authentication (I&#39;m poor but I sweat node.js Why the library on the partnership ..." [ja] -- こば@変人生徒会長. http://twitter.com/KOBA789/status/21333825794
[18:00] creationix: does JS's RegExp support named capture groups?
[18:00] tmpvar has joined the channel
[18:01] herbySk has joined the channel
[18:01] rnewson: but if (req.writable) { req.write() } actually helps. Are you saying it shouldn't? :)
[18:01] mscdex: creationix: no, at least according to http://www.regular-expressions.info/javascript.html
[18:01] rnewson: I'm not optimizing for the failed write case. I'm testing couchdb so I don't expect anything but 201 all the way. it's just if the database isn't there that the test actually crashes, rather than give a report full of 404's.
[18:01] _announcer: Twitter: "Why do I bother to get an node.js? Well, I&#39;m from the Javascript that I love. Inspired by the announcement that my graduate research Javascripter The more I think that one person" [ja] -- こば@変人生徒会長. http://twitter.com/KOBA789/status/21333911701
[18:01] creationix: mscdex: yeah, I don't trust that site though
[18:02] mscdex: creationix: well, then the only other way to know is to try and see if it breaks
[18:02] mscdex: heh
[18:02] creationix: mscdex: sure, I'm just trying to see what the syntax would be if it's supported
[18:02] mscdex: rnewson: yes, check req.writeable before doing .write()
[18:03] rnewson: mscdex: there's a small race there, right? I'm fine with that, just checking.
[18:03] creationix: rnewson: what is the race?
[18:04] rnewson: that it becomes not writable after I check and before I write?
[18:04] creationix: mscdex: I do trust this site https://developer.mozilla.org/en/JavaScript/Guide/Regular_Expressions#Using_Parenthesized_Substring_Matches and it doesn't mention named capture groups
[18:04] creationix: rnewson: how would that happen, it's single threaded
[18:04] _announcer: Twitter: "Finally a #NodeJS #MySQL driver I can trust http://bit.ly/ccBhrd #persistencejs" -- Zef Hemel. http://twitter.com/zef/status/21334095045
[18:04] rnewson: creationix: .writable is not a property of the underlying socket?
[18:04] jamescarr_: darn.... brain block today
[18:04] hansek has joined the channel
[18:05] TangoIII has joined the channel
[18:05] mscdex: creationix: here you go: http://xregexp.com/
[18:05] mscdex: :p
[18:05] mscdex: rnewson: it's .writeable
[18:05] creationix: rnewson: hmm, not sure about that, but I think it's safe
[18:06] _announcer: Twitter: "@jkreeftmeijer Do you feel node.js & rails / sinatra / ramaze / merb (or whatever) mesh pretty well?" -- Paul Nakata. http://twitter.com/icey/status/21334176249
[18:06] creationix: node does use threads internally, so I'm not ruling out the possibility, but I think that's mostly for file I/O
[18:06] creationix: mscdex: thanks, but I think I'll just use numbered matches
[18:06] paul__ has joined the channel
[18:06] rnewson: creationix: can't see how it can not have a race condition, but if req.writeable is false when it's seen a 4xx response, that's good enough for me.
[18:07] rnewson: creationix: but it can't be blocking for the statusCode.
[18:09] _announcer: Twitter: "if u have problem with unassigned associations in sequelize, check out the latest commit on github #nodejs #orm #mysql" -- Sascha Depold. http://twitter.com/sdepold/status/21334367662
[18:09] piotrekm: mscdex: you sure? I have it '.writable' http://pastebin.com/pBCEhXbt
[18:10] jochen1 has joined the channel
[18:10] daleharvey has joined the channel
[18:11] rnewson: mscdex: it's .writable in http.js and .writeable in fs.js. how fun!
[18:11] joshbuddy has joined the channel
[18:12] qFox has joined the channel
[18:12] Sembiance: afternoon folks :)
[18:12] mscdex: Sembiance: no, i'm more like a 24/7 folk
[18:12] mscdex: :p
[18:13] matt_c: Hello intarnets!
[18:14] wllm has joined the channel
[18:14] mikeal has joined the channel
[18:15] _announcer: Twitter: "Just released async.js v0.1 for node.js http://groups.google.com/group/nodejs/browse_thread/thread/ade9fffbaca49a44#" -- Fabian Jakobs. http://twitter.com/fjakobs/status/21334726716
[18:16] Sembiance: mscdex: :)
[18:19] bradleymeck: anyone think __noSuchMethod__ is really needed vs proxies?
[18:20] jochen1: ah yes __noSuchMethod__ would make this so much easier
[18:20] jochen1: *things
[18:21] bradleymeck: but the question i ask is "for what use case"
[18:21] rsms has joined the channel
[18:22] jochen1: lazy loading stuff :)
[18:22] bradleymeck: why not check if its not there instead of using __noSuchMethod__?
[18:22] jochen1: say I want to do: Dialog.open() but dialog.js with all its methods is not yet loaded
[18:23] Rixius has joined the channel
[18:23] jochen1: then I can catch that, load the script and then execute the method
[18:23] bradleymeck: but you can already do that
[18:24] jochen1: ? whatever… just want to show you that some people do appreciate __noSuchMethod__ :)
[18:24] bradleymeck: i dont go against it, i just dont see the actual use case :/
[18:25] dnm: Hi all. Here's a potentially stupid question, apologies in advance. I just found out about Node.js last night/today. It seems to be ((V8) + (bindings for asynchronous I/O OS APIs)). Is that it? Is there anything else I'm missing? I know it also has an implicit event loop, but outside of that, is there a big piece (or pieces) I'm not seeing? Thanks.
[18:25] _announcer: Twitter: "what people don't get about node.js/eventmachine: CPU bound tasks must be moved into separate processes. otherwise your QOS will suck." -- stkaes. http://twitter.com/stkaes/status/21335363179
[18:25] _announcer: Twitter: "if anyone wants to join the node_knockout @ August 28th, my team still has two vacancies # nodejs" [pt] -- Fabricio Zuardi. http://twitter.com/fczuardi/status/21335365231
[18:25] mattly has joined the channel
[18:26] maushu has joined the channel
[18:26] bradleymeck: dnm, thats the gist of it, there is good support for various web protocols lying about including ssl, a nice binary array with Buffers, and focus on streaming rather than buffering
[18:27] dnm: bradleymeck: Thanks!
[18:28] rwaldron has joined the channel
[18:28] mu-hannibal: anyone else having problems installing node-glob? getting this:
[18:28] mu-hannibal: ../src/glob.cc: In function `int EIO_GlobAfter(eio_req*)':
[18:28] mu-hannibal: ../src/glob.cc:100: error: `free' was not declared in this scope
[18:28] mu-hannibal: ../src/glob.cc: In function `v8::Handle<v8::Value> GlobAsync(const v8::Arguments&)':
[18:28] mu-hannibal: ../src/glob.cc:117: error: `calloc' was not declared in this scope
[18:29] JimBastard: jochen1: you have an event fire when your dialog.js loads, you shouldn't be doing a method missing for that.....at all
[18:29] aubergine has joined the channel
[18:30] jetienne has joined the channel
[18:30] jochen1: whatever, it works for me :)
[18:31] JimBastard: what are you polling?
[18:31] JimBastard: if typeof foo != 'undefined'
[18:33] bradleymeck: well the ability to see what you were throwing as arguments used to be an ability to lazy load in complex things with options, but Proxies are the new way ecma ppl want it, which is fine by me. I think there has to be a good edge case noSuchMethod covers, but i cant think of it
[18:33] mu-hannibal has left the channel
[18:35] stagas has joined the channel
[18:35] bradleymeck: cause proxies are heavier than __noSuchMethod__ but they remove a test whenever you access an object that __noSuchMethod__ uses, so its probably equal in time
[18:37] Rixius has joined the channel
[18:38] _announcer: Twitter: "♻ @ lucabastos: In the talk on the CEP # TDC2010 summer toast Node.JS&#39;ll show a bit of a CEP engine with Javascript" [pt] -- diogobaeder. http://twitter.com/diogobaeder/status/21336147500
[18:41] jamescarr_: man... I always have to re-edit articles
[18:41] jamescarr_: I use too many we, you, I, etc
[18:44] behmann has joined the channel
[18:44] paul__ has joined the channel
[18:45] femtoo has joined the channel
[18:46] _announcer: Twitter: "Seeing work like this http://bit.ly/cvRIIT (mysql for Node.js) is one of the most exciting things I can imagine for web dev today. #MOAR" -- Luke Smith. http://twitter.com/ls_n/status/21336630682
[18:48] justin___ has joined the channel
[18:51] malkomalko has joined the channel
[18:52] bradleymeck: haha more pronouns, it and they give +2!
[18:52] dnm: Bye for now!
[18:52] dnm has left the channel
[18:53] teemow has joined the channel
[18:53] jsilver has joined the channel
[18:54] saikat has joined the channel
[18:54] jakehow has joined the channel
[18:56] jherdman has joined the channel
[18:56] jochen1 has joined the channel
[18:58] mu-hannibal has joined the channel
[18:58] exlibris has joined the channel
[18:59] matt_c__ has joined the channel
[19:01] blogometer: Checking out Node.js on Windows XP to build it eio.c with Visual Studio 2008 C++.
[19:01] blogometer: And see what's missing.
[19:01] mu-hannibal: q: how would you read a property file during startup? Synchronous is fine...
[19:03] bradleymeck: property file in what format? fs.readFileSync is there too
[19:04] bradleymeck: how do ppl in here feel about the argument to start using some STL stuff in node's core?
[19:04] mu-hannibal: just a list of directory names separated by newlines 
[19:04] mbleigh has joined the channel
[19:05] zapnap has joined the channel
[19:06] konobi: a node/websockets version of this would be awesome... http://code.google.com/p/androidome/
[19:06] mu-hannibal: bradleymeck: thx - just somehow have a hard time parsing the API docs - is there already a tips/tricks site?
[19:08] _announcer: Twitter: "MySQL adapter for Node.js, aka JS wearing its grown up pants. http://bit.ly/9a08pl (tip: @ls_n)" -- Micah Snyder. http://twitter.com/dfltr/status/21337940677
[19:09] bradleymeck: require("fs").readFileSync("list.txt").toString().split("\n").forEach(function(line,i){console.log(i+">>"+line)})
[19:09] mu-hannibal: nice - will add to my node.js cheatsheet
[19:12] charlenopires has joined the channel
[19:15] Astro has joined the channel
[19:16] kersny has joined the channel
[19:20] _announcer: Twitter: "Hummingbird - Millisecond Real-time Web Analytics with Node.js http://bit.ly/aG0rOu #design #frameworks" -- Jessyka. http://twitter.com/yummymashup/status/21338618651
[19:22] Tobsn has joined the channel
[19:23] mattly has joined the channel
[19:25] jochen1: anyone got an idea what this means? =>
[19:25] jochen1: Build failed:  -> task failed (err #1): 
[19:25] jochen1: 	{task: cxx platform_linux.cc -> platform_linux_4.o}
[19:25] jochen1: while trying to install node.js on CentOS release 4.8
[19:25] jochen1: 2.6.9-89.0.25.ELsmp
[19:27] icy: that means it failed to compile platform_linux.cc
[19:27] jochen1: `PR_SET_NAME' was not declared in this scope meh I just commented it in src/platform_linux.cc :)
[19:27] icy: try with -v
[19:29] wllm has joined the channel
[19:33] EyePulp: love doing a git pull and seeing so many changes come down the wire.
[19:33] hojberg has joined the channel
[19:33] wattz: this meeting is making kittens cry
[19:36] bradleymeck: do their tears cure cancer?
[19:37] stride: you should yell something in french
[19:37] bradleymeck: je suis un chou chou avec fondue!
[19:37] scoates: C'est nous qui voulons les craquelins!
[19:37] huyhong has joined the channel
[19:40] feroz_ has joined the channel
[19:40] sideshowcoder has joined the channel
[19:43] Gruni has joined the channel
[19:47] danielzilla has joined the channel
[19:47] mattly has joined the channel
[19:53] charlesjolley- has joined the channel
[19:55] _announcer: Twitter: "Hummingbird real time traffic visualization in Node.js / JavaScript / WebSockets http://is.gd/ekxNt #weblabor" -- Peter Galiba. http://twitter.com/Poetro/status/21340635913
[19:55] FransWillem: w00t!
[19:55] bradleymeck: :?
[19:55] FransWillem: Got a simple web server proxy to newsgroups :p
[19:56] FransWillem: Allows me to view yEnc encoded JPGs, and browse newsgroups :p
[19:57] FransWillem: At least, single part yenc files :p
[19:59] drudge: ew :P
[19:59] mattly_ has joined the channel
[20:00] _announcer: Twitter: "Something tells me Node.js is going to be something to be reckoned with in the near future... just a hunch" -- Sean Nieuwoudt . http://twitter.com/SeanNieuwoudt/status/21340922691
[20:01] MrTopf has joined the channel
[20:02] bradleymeck: !tweet @SeanNieuwoudt ... lies ... 
[20:02] jsilver has joined the channel
[20:02] _announcer: Twitter: "@stkaes What you meant to say: What people that shouldn't develop anything but CRUD applications don't get about node.js/eventmachine" -- Martin S. http://twitter.com/MSch/status/21341068365
[20:04] bradleymeck: that msg confuses me
[20:04] drudge: i don't understand what that doesn't not say
[20:05] paulwe has joined the channel
[20:05] sideshowcoder has joined the channel
[20:08] saikat has joined the channel
[20:09] rsms has joined the channel
[20:10] bradleymeck: what ppl that shouldnt develop... (what is "that")
[20:11] stride: simple webapps that only retrieve & present stuff with mostly trivial business logic
[20:11] Ori_P has joined the channel
[20:12] astrolin has joined the channel
[20:13] joshbuddy has joined the channel
[20:15] maushu has joined the channel
[20:24] rnewson has joined the channel
[20:28] _announcer: Twitter: "Sunday project: memebot, a pluggable IRC bot in CoffeeScript on node.js @ http://is.gd/ekA0n" -- Martin Häger. http://twitter.com/mtah/status/21342535991
[20:30] jackish has joined the channel
[20:30] _announcer: Twitter: "asciimo - create awesome ascii art with javascript! works in the browser or node.js http://goo.gl/fb/5yCRQ" -- faboulousGR. http://twitter.com/faboulousGR/status/21342665255
[20:34] tilgovi has joined the channel
[20:37] javajunky has joined the channel
[20:38] mbleigh has joined the channel
[20:40] joshbuddy has joined the channel
[20:40] joshbuddy has joined the channel
[20:42] _announcer: Twitter: "node-mail: send email safely with Node.JS (supports gmail) http://j.mp/bzznwD #nodejs" -- Ben Weaver. http://twitter.com/bwvr/status/21343328447
[20:43] zomgbie has joined the channel
[20:44] TrueShot_ has joined the channel
[20:44] dannycoates has joined the channel
[20:45] matt_c has joined the channel
[20:47] Gruni has joined the channel
[20:49] JimBastard has joined the channel
[20:49] javajunky has joined the channel
[20:51] digitalspaghetti: Another SMTP library *sigh*
[20:52] mape: The more the merrier
[20:52] digitalspaghetti: mind you this one looks a lot cleaner
[20:52] voodootikigod has joined the channel
[20:52] Ori_P has joined the channel
[20:53] JimBastard: digitalspaghetti: the more people who want to re-implement the same shit, the better
[20:53] JimBastard: if people start to complain there are a lot of shit apps in a certain domain, i'll be the first person to yell as loud as i can to not use the shitty ones
[20:53] JimBastard: and i'll pick a winner :-)
[20:53] bradleymeck: i just want me some attachment action
[20:53] digitalspaghetti: smtp does seem to be the flavor of the week :)
[20:53] JimBastard: bradleymeck: i have attachment support actually
[20:53] JimBastard: i havent pulled yet
[20:55] JimBastard: bradleymeck: http://github.com/Brucer/node_mailer/commit/bb53c771304ebaf8802e5dee669eba283b8d6ab5
[20:55] mape: Hmm people really need to add their things to npm, I've been noticing I won't try things that aren't on there
[20:55] guybrush: 
ACTION nods

[20:56] javajunky: mape: some stuff that is there doesn't work :( 
[20:56] digitalspaghetti: JimBastard: looks like it would be worth adding to this one then http://github.com/weaver/node-mail
[20:56] javajunky: connect springs to mind ;) 
[20:56] bradleymeck: oooo
[20:56] digitalspaghetti: this one has a nice flow to it
[20:56] JimBastard: digitalspaghetti: i thought thats what we were talking about
[20:56] JimBastard: digitalspaghetti: i dont think it does
[20:56] javajunky: I've finally ported christkv's oauth provider to connect, connect now has oauth provider capabilities for free, yay go open source :)
[20:57] bradleymeck: still havent seen anyone try the github npm push hook
[20:57] paul__ has joined the channel
[20:58] JimBastard: actually digitalspaghetti i take that back, it has a few nice things
[20:59] daleharvey has joined the channel
[21:00] Tim_Smart has joined the channel
[21:02] ajpi has joined the channel
[21:04] FransWillem: creationix: Why did you choose to have a function(callback,errback) style for Do instead of the Node.js default function(err,data) callback ?
[21:04] creationix: FransWillem: because the node style didn't exist at the time
[21:04] femtooo has joined the channel
[21:04] sideshowcoder has joined the channel
[21:04] creationix: and it was thought that nobody would use raw callbacks
[21:05] creationix: so Do was competition to Promises
[21:05] creationix: FransWillem: I use Step now, I like the pure callback approach
[21:06] FransWillem: Ah, ok
[21:06] FransWillem: and Step also handles non-async callbacks, right ?
[21:06] creationix: FransWillem: I believe so, if not it's a bug
[21:06] FransWillem: k, thanks :)
[21:06] creationix: FransWillem: I do know Step supports return values and exceptions
[21:06] creationix: it forwards them to the next step's callback
[21:07] FransWillem: Shame Do is no longer used though
[21:07] FransWillem: I kinda liked the whole specify as steps
[21:07] FransWillem: instead of Step's chained callback functions
[21:07] creationix: I think it's used, I just don't use it
[21:07] tav_ has joined the channel
[21:07] creationix: FransWillem: have you looked at conductor?
[21:07] creationix: it's very declarative
[21:08] FransWillem: Lol, how many libraries like that did you write :p?
[21:08] FransWillem: Hmm, can't find any examples for conductor
[21:09] FransWillem: nor docs
[21:09] ph^ has joined the channel
[21:10] JimBastard: creationix really is just the front man for a team of developers
[21:10] JimBastard: :-)
[21:10] creationix: JimBastard: that explains my multiple personality disorder ;)
[21:10] creationix: FransWillem: http://github.com/creationix/conductor/blob/master/test/test.js
[21:10] satori_ has joined the channel
[21:10] JimBastard: i thought that was marak who has the mpd
[21:10] nerdEd has joined the channel
[21:10] creationix: FransWillem: http://howtonode.org/step-of-conductor
[21:10] JimBastard: i actually have 4 personalities, all javascript related 
[21:10] creationix: FransWillem: I think that's all the docs
[21:11] JimBastard: tmpvar has been making great progress on the UI front for conductor 
[21:11] jakehow has joined the channel
[21:11] creationix: JimBastard: exciting
[21:11] creationix: JimBastard: did you know I made conductor for him as a favor, and then he paid me!  That's awesome
[21:12] FransWillem: Wow, apart from the horrible declaration syntax, conductor looks uber nice
[21:12] JimBastard: hee hee 
[21:12] nrstott has joined the channel
[21:12] creationix: FransWillem: yeah, and with the UI tmpvar is working on, the ugly syntax won't matter
[21:12] JimBastard: FransWillem: the goal is to build a UI for it ....in the browser...using canvas....
[21:12] JimBastard: with drag and drops
[21:12] FransWillem: Ah, neat
[21:13] JimBastard: its really exciting stuff
[21:13] JimBastard: visual programming for node
[21:13] JimBastard: will REALLY help people grok async
[21:13] mikeal has joined the channel
[21:13] jherdman has joined the channel
[21:14] codemariner has joined the channel
[21:14] FransWillem: Hmm, is there any way to have an output that should be re-evaluated each time it's used?
[21:14] creationix: ?
[21:14] FransWillem: Well
[21:15] FransWillem: I'm writing a simple NNTP indexer
[21:15] FransWillem: and sometimes when I don't do anything
[21:15] FransWillem: the server will disconnect me for being idle
[21:15] mscdex: probably have to send a NOOP
[21:15] mscdex: or something
[21:15] creationix: FransWillem: object getter?
[21:15] creationix: not sure what you're looking for
[21:16] FransWillem: Something to handle these kind of functions: http://pastebin.com/VMw9xEUL
[21:16] creationix: a = {get time() { return Date.now(); }}; a.time; a.time
[21:16] FransWillem: creationix: Yes, but then async :p
[21:16] creationix: interesting, haven't thought about that
[21:17] creationix: FransWillem: I have an async cache and queue pattern I use a lot
[21:17] creationix: FransWillem: http://github.com/creationix/node-git/blob/master/lib/git.js#L57-99
[21:17] creationix: it's a "decorator" for async functions
[21:17] creationix: if the same function is called twice in a row with the same arguments, then the action is only performed once
[21:18] wattz: i have missed you all today
[21:18] creationix: wattz: welcome back :)
[21:18] FransWillem: Oh well, guess I'll stick with Step :)
[21:18] creationix: 
ACTION goes back to rewriting nStore to be awesome

[21:18] creationix: FransWillem: cool
[21:18] wattz: thank you ;)
[21:18] wattz: google's protobufs look cool
[21:18] saikat has joined the channel
[21:19] creationix: wattz: have you seen msgpack?
[21:19] creationix: http://msgpack.org/
[21:19] creationix: same idea, different format
[21:19] wattz: cool
[21:19] wattz: looking
[21:19] zemanel has joined the channel
[21:19] wattz: ruby?
[21:19] FransWillem: creationix: Don't suppose you've ever thought of writing a front-end to nStore that uses the same API as redis-node-client? That way people can work with nStore normally, then if somehow nStore isn't scaling anymore, can simply drop in redis.
[21:20] wattz: ohhh all langs
[21:20] creationix: wattz: node here http://github.com/pgriess/node-msgpack
[21:20] creationix: FransWillem: no, but it shouldn't be hard for someone to wrap nstore to do just that
[21:20] creationix: FransWillem: I'm more worried about making nStore never lose data right now
[21:21] wattz: creationix: i was thinking of throwing together a quick datastore in C/C++ from protobufs for project at work
[21:21] creationix: wattz: that would be fun
[21:21] wattz: yeah
[21:21] wattz: but i have to finish the last piece of Bob before i can start :P
[21:22] wattz: with protobufs i would basically be writing an embeddable store
[21:25] dnm has joined the channel
[21:25] dnm: Back!
[21:25] wattz: creationix: do you develop in C++?
[21:25] creationix: wattz: no, I learned it 10 years ago, never used it since
[21:25] wattz: ahh
[21:26] creationix: you'll notice all my libraries (including my postgres implementation) are pure js
[21:26] wattz: ahh
[21:26] wink_: man thats froggy as hell :p
[21:26] wattz: i have to say i have enjoyed relearning/expanding my knowledge of C++
[21:26] wattz: writing stuff for node and v8
[21:26] creationix: I prefer C to C++
[21:26] wattz: ahh
[21:26] wattz: Im learning lex/yacc now
[21:26] creationix: but yeah, I'm not the best at either
[21:26] wattz: creationix: have you seen Bob? :D
[21:27] creationix: I think you showed me
[21:27] wattz: Make + Configure (with a touch of Waf)
[21:27] wattz: standalone
[21:27] creationix: neat
[21:27] wattz: http://github.com/wess/Bob-The-Builder
[21:27] wattz: i got configure working, i just have to put the build component in
[21:27] maushu: wth
[21:27] wattz: for prelim testing
[21:27] maushu: Die in a fire.
[21:28] wattz: ._o
[21:28] wattz: maushu: me?
[21:28] maushu: Who the hell names a project like that?
[21:28] wattz: my 3 year old.
[21:28] wink_: lol
[21:28] wattz: I asked him what I should name "a program that builds programs"
[21:28] maushu: Don't let your 3 year old name projects.
[21:28] wattz: he said
[21:28] wattz: "Bob the Builder"
[21:29] charlesjolley- has joined the channel
[21:30] wattz: fine i will rename it mouthToDick so people can say "Im going to use mouthToDick"
[21:30] wattz: :D
[21:31] daniellindsley2: JimBastard: Just wanted to say thanks. JSONloops is an unending source of joy.
[21:35] jherdman has joined the channel
[21:36] JimBastard: lol really daniellindsley2 ?
[21:36] JimBastard: my computer is too slow with afplay to keep a beat
[21:36] JimBastard: ive been begging for a proper audio out channel for months
[21:37] daniellindsley2: Yes really. Not using it for anything serious but I like it just the same.
[21:39] JimBastard: hee hee thanks daniellindsley2 
[21:39] jamescarr_: hey, anyone here know how to enable ssl for node.js?
[21:39] wattz: aiight, back to lex/yacc
[21:39] JimBastard: i like making silly experiments :-)
[21:39] jamescarr_: aside from running on port 443?
[21:39] wattz: JimBastard: you are my silly experiment ;)
[21:39] ryan_gahl has joined the channel
[21:40] bradleymeck: jamescarr_ whats the problem?
[21:41] jamescarr_: bradleymeck, didnt find any documentation on it
[21:41] jamescarr_: thats all
[21:41] bradleymeck: theres some, all there is to it is setting the secure argument of various things to true
[21:41] JimBastard: There he goes. One of God's own prototypes. A high-powered mutant of some kind never even considered for mass production. Too weird to live, and too rare to die. 
[21:42] wllm has joined the channel
[21:43] blogometer: Is asynchronous mmap a thing?
[21:43] blogometer: Or did I just make it up?
[21:43] mau2 has joined the channel
[21:43] atmos has joined the channel
[21:44] wattz:  lol
[21:45] blogometer: Stop laughing.
[21:45] blogometer: Does Node.js do any buffering? If I keep reading the same section of a file using Node.js, will the right things happen? Or do I need mmap? If so, can you do mmap in an async way?
[21:47] FransWillem: blogometer: mmap is, by definition, not asynchronous.
[21:47] blogometer: It is by definition synchronous.
[21:47] felixge has joined the channel
[21:47] blogometer: Therefore.
[21:48] blogometer: Seems like, there ought to be away to ensure that, if there is a page fault, the read can be done async.
[21:48] FransWillem: blogometer: Basically what it does is tell the OS "give me some memory, and when I read it, I want it to be what's in that file. If I read from it, and you don't actually have the file in memory yet, just block me for a while"
[21:48] marshall_law has joined the channel
[21:48] FransWillem: blogometer: Well, that'd be just a simple cache.
[21:48] FransWillem: e.g. if (cache) return cache; else somehow get the cache
[21:49] FransWillem: something like this http://pastebin.com/i1WMR9cq (warning: horrible code ahead)
[21:49] blogometer: FransWillem: read from file: yes. So, I'm implementing a B-tree, and I'm wondering if ...
[21:49] blogometer: FransWillem: Okay. That is some form of caching for Node.js?
[21:49] jamescarr_: what are those compliance rules for handling online payments? PCI? LCI?
[21:50] drudge: PCI jamescarr_
[21:50] FransWillem: blogometer: You'll have to implement your own caching
[21:51] marshall_law_ has joined the channel
[21:51] blogometer: FransWillem: It seems that, mmap would be a great way to do it, if it where not synchronous only, caching is not as good, more complex, but async.
[21:51] aubergine has joined the channel
[21:52] jamescarr_: having an awesome day wrestling with maven
[21:53] blogometer: Maven is Satan's spawn with angle brackets.
[21:53] FransWillem: blogometer: How is mmap better than caching manually ?
[21:54] javajunky: *sob* is npm dead atm ?
[21:54] mape: the repo seems to spit json at least
[21:54] mape: javajunky: works here
[21:55] javajunky: mape: cheers, eird, I'm getting DNS errors
[21:55] javajunky: whats the host  ?
[21:55] blogometer: mmap is a form of caching, it is kind of done and I don't have to think about it.
[21:56] javajunky: mape: and now it workss…weird !
[21:56] mape: hehe
[21:56] Blackguard has joined the channel
[21:57] jamescarr_: authorizenet is causing me problems :(
[21:57] JimBastard: jamescarr_: bad https client requests?
[21:57] jamescarr_: javajunky, I had to turn ssl off for npm to publish
[21:57] bradleymeck: FransWillem mmap is only better when you are taking up so much you know its going to go into swap space
[21:58] JimBastard: jamescarr_: we are experiencing issues doing ssl to rackspace atm....only GET working...no PUT or POST
[21:58] jamescarr_: JimBastard, connection reset when running integration specs
[21:58] jamescarr_: JimBastard, it's annoying really... nothing changed since yesterday
[21:58] jamescarr_: leading me to believe it's an external influence
[22:00] JimBastard: jamescarr_: hrmmmm
[22:01] javajunky: ooo reflog never used that before, git continues to surprise.
[22:01] JimBastard: well keep debugging on it, if you hit a complete wall we might be able to help out. very backlogged this month :-)
[22:01] Tim_Smart: Hmm C pointers are confusing me. I think I'll get there eventually.
[22:01] JimBastard: which is a good thing
[22:02] aubergine has joined the channel
[22:04] bradleymeck: C pointers?
[22:05] jakehow has joined the channel
[22:06] kodisha has joined the channel
[22:06] Wandrewvious has joined the channel
[22:07] mw_ has joined the channel
[22:10] alexkwolfe has joined the channel
[22:11] derferman has joined the channel
[22:12] paul__ has joined the channel
[22:15] Tim_Smart: bradleymeck: http://cplus.about.com/od/learningc/ss/pointers.htm
[22:15] dnolen has joined the channel
[22:16] bradleymeck: aaah, yea, pointers are a bit odd to think about
[22:17] unomi has joined the channel
[22:19] admc has joined the channel
[22:21] ryah: strange - rabbitmq has this application layer flow control thing...
[22:22] mape: Does it use magic?
[22:23] JimBastard: magicmq
[22:24] jamescarr_: rabbitmq rocks
[22:26] JimBastard: we use it at work to pass messages between services
[22:26] sonnym has joined the channel
[22:27] Dmitry1 has joined the channel
[22:27] twoism has joined the channel
[22:29] maushu has joined the channel
[22:29] bsstoner has joined the channel
[22:30] jamescarr_: JimBastard, dont use it at work, we use Oracle Weblogic's crap JMS stuff
[22:30] jamescarr_: trying to get them to use RabbitMQ though
[22:31] jamescarr_: need to find someway to escape the day job :(
[22:31] jamescarr_: my day job vs. side hacking is kind of like a Bruce Wayne / Batman setup
[22:31] jamescarr_: Bruce Wayne is boring but pays the bills
[22:31] mw_: except that you'd have a hard time holding your own in a fight against superman :)
[22:32] mw_: (i assume)
[22:32] jamescarr_: mw_, not really... after all Batman owned Superman in The Dark Knight Returns
[22:32] mw_: that's what i was thinking of, and not many people can do that
[22:32] mw_: but batman could
[22:34] ajpiano has joined the channel
[22:36] jamescarr_: so back to my earlier question... how do you setup expressjs to use ssl?
[22:36] jamescarr_: just listen on port 443? how do you setup the site ssl cert?
[22:38] mattly has joined the channel
[22:42] rauchg_: jakehow: 
[22:42] rauchg_: jamescarr_: 
[22:42] rauchg_: use spark
[22:42] rauchg_: to start your express app
[22:42] jamescarr_: oh?
[22:42] Astro: why is Buffer::New() deliberately private?
[22:43] ajpi has joined the channel
[22:43] sveisvei has joined the channel
[22:45] bradleymeck: im seeing it as public?
[22:46] huyhong has left the channel
[22:47] kodisha has joined the channel
[22:48] Astro: bradleymeck: uh, no?
[22:48] Astro: bradleymeck: node_buffer.h?
[22:48] Astro: HEAD = ae5d613?
[22:49] bradleymeck: line 36?
[22:49] shimondoodkin has joined the channel
[22:49] mbleigh has joined the channel
[22:51] samsonjs has joined the channel
[22:51] shimondoodkin: i have a problem with http if i define the on end and on data callbacks after like 120 ms it does not call callbacks or maybe if it is clouse inside  a clousre, the object looks ok but it does not fire the events
[22:52] bradleymeck: is the request done by 120ms?
[22:52] shimondoodkin: it hangs
[22:52] joshbuddy has joined the channel
[22:53] shimondoodkin: after a minute it fails
[22:55] jsilver has joined the channel
[22:55] daleharvey has left the channel
[23:03] stepheneb has joined the channel
[23:04] gwoo has joined the channel
[23:05] tpryme has joined the channel
[23:09] ink83_ has joined the channel
[23:10] keeto has joined the channel
[23:11] blogometer has joined the channel
[23:12] zapnap has joined the channel
[23:18] Astro: bradleymeck: sorry, I meant the one that returns a v8::Handle<v8::Value>
[23:18] mtodd has joined the channel
[23:18] devinus has left the channel
[23:23] technoweenie has joined the channel
[23:23] bradleymeck has left the channel
[23:23] falconair has joined the channel
[23:26] bsstoner has left the channel
[23:28] maqr has joined the channel
[23:30] tekky has joined the channel
[23:33] streampunk has joined the channel
[23:35] shimondoodkin: http://gist.github.com/527981
[23:36] shimondoodkin: bradleymeck: do you want to test?
[23:36] Throlkim has joined the channel
[23:36] sh1mmer: technoweenie: yo
[23:37] sh1mmer: is the current head of riak-js stable(ish)
[23:38] jashkenas has joined the channel
[23:40] blogometer: Kind of an async mmap:
[23:40] blogometer: http://code.google.com/p/msolo/wiki/LighttpdAIO
[23:41] creationix: is anybody using nStore for fast inserts?
[23:42] creationix: I'm trying to rewrite it to be more robust and I'm only getting 15k inserts/second (inserting 1,000,000 unique records)
[23:42] dannycoates has joined the channel
[23:42] shimondoodkin: are you doing console.log?
[23:43] stepheneb has joined the channel
[23:43] inimino: Is npm publish currently not working?
[23:44] tilgovi has joined the channel
[23:44] maushu: Dammit, I need a deconstructor.
[23:45] shimondoodkin: maushu: ther eis an on exit event
[23:45] maushu: For objects?
[23:45] shimondoodkin: no
[23:45] streampunk: maushu: you could try RAII style management.
[23:46] shimondoodkin: maushu: you may do a function like destruct
[23:46] streampunk: That would make destruction calls explicit but it might strengthen your code anyway.
[23:46] maushu: Already have one called dispose.
[23:46] jashkenas has left the channel
[23:46] jashkenas has joined the channel
[23:46] maushu: But if I lose the reference to the object memory will leak.
[23:47] maushu: Well, more like resources will leak.
[23:47] streampunk: maushu: Right. That's why RAII is useful for deciding who should keep that last reference.
[23:48] shimondoodkin: is the idea is to have a different function that passes thru each time you create an object?
[23:48] streampunk: It's more so for direct reference languages like C++ but it can help for GC'ed languages too when resource holding becomes a complex situation.
[23:49] streampunk: It basically ties object life to the scope of something else like the stack or a different object. Always the same thing that allocates it.
[23:49] shimondoodkin: and you have an array of objects you have created
[23:49] maushu: streampunk, doesn't it still need a deconstructor?
[23:49] streampunk: RAII is simple so the fact it has an acronym is almost overkill. 
[23:49] streampunk: maushu: You can use an explicit call in this case. Just a try finally pair.
[23:51] maushu: Hmm.
[23:51] JimBastard has joined the channel
[23:52] streampunk: I'll have to write something on this later. Still finishing up notes from my workshop that I'll post.
[23:53] shimondoodkin: streampunk: maybe you have an idea why i have a bug in this: http://gist.github.com/527981
[23:54] JimBastard: streampunk: you trying out javascript
[23:54] JimBastard: ?
[23:54] JimBastard: its great
[23:55] streampunk: JimBastard: It's just like BASIC if you hold down shift. I feel right at home. :P
[23:55] hij1nx has joined the channel
[23:55] streampunk: shimondoodkin: looking.
[23:56] [[zz]] has joined the channel
[23:56] streampunk: shimondoodkin: I'd recommend moving {'s to the same line as the function() btw... just avoids odd silent syntax holes in JS.
[23:57] charlenopires has joined the channel
[23:58] maushu: streampunk, lulz
[23:58] maushu: BASIC is better, it has GOTO.
[23:59] streampunk: Yeah. JS might make it there one day.
[23:59] streampunk: Here's hoping. =D
[23:59] streampunk: PHP even has goto now.
[23:59] vdrab has joined the channel
[23:59] JimBastard: we have goto
[23:59] ryan2 has joined the channel
[23:59] shimondoodkin: for me thats ok, i use to it, it is useful in text only editors without hilighting it helps me solve barakets problems quickly without any sophisticated tools