[00:00] c4milo has joined the channel
[00:00] jakehow has joined the channel
[00:05] dpritchett: wow i really enjoyed this writeup http://www.catonmat.net/blog/node-js-knockout-competition
[00:05] dpritchett: three people in different countries building a node knockout app in 48 hours using irc and github
[00:06] JimBastard_: YSBAT
[00:07] dpritchett: 
ACTION googles frantically

[00:09] JimBastard_: ohh yeah no, i made that one up
[00:09] JimBastard_: You Should Blog About That
[00:09] JimBastard_: :-)
[00:10] JimBastard_: i have a hand carved wooden sign on my wall that says that
[00:10] bruse: you should make it virtual so that dpritchett can find it on google
[00:11] dpritchett: i finally figured it out i think
[00:11] mr_daniel has joined the channel
[00:12] dpritchett: oh, i found "you should be able to" after a few minutes' googling
[00:12] [[zz]] has joined the channel
[00:12] JimBastard_: lolwut
[00:12] lstoll: dpritchett: we did the same, using campfire, xmpp and skype instead of IRC
[00:13] lstoll: I should run the numbers over that
[00:13] dpritchett: that sounds like fun
[00:13] lstoll: the skype part definitely helped though. Arguing in text is nothing compared to arguing with voice.
[00:14] mjr_ has joined the channel
[00:17] c4milo has joined the channel
[00:19] alcuadrado has joined the channel
[00:19] c4milo1 has joined the channel
[00:20] c4milo1 has joined the channel
[00:26] alexb_ has joined the channel
[00:26] c4milo has joined the channel
[00:31] jakehow has joined the channel
[00:33] twomashi has joined the channel
[00:33] twomashi has left the channel
[00:33] AAA_awright: lstoll: What's Campfire?
[00:34] lstoll: AAA_awright: http://campfirenow.com/
[00:34] c4milo has joined the channel
[00:34] AAA_awright: Ah
[00:37] muk_mb has joined the channel
[00:40] evanpro has joined the channel
[00:41] deepthawtz has joined the channel
[00:46] wink_ has joined the channel
[00:47] alexb_ has joined the channel
[00:53] bencc has left the channel
[00:56] jamescarr_: fuck
[00:56] stephenjudkins_ has joined the channel
[01:01] [[zz]] has joined the channel
[01:01] ekidd has joined the channel
[01:05] secoif has joined the channel
[01:05] TheLastBookworm has joined the channel
[01:07] TheLastBookworm: hello
[01:10] threeve has joined the channel
[01:11] tilgovi has joined the channel
[01:13] Tim_Smart has joined the channel
[01:18] Blink7 has joined the channel
[01:19] danielzilla has joined the channel
[01:26] melpad has joined the channel
[01:39] itissid has joined the channel
[01:40] itissid: Ho i was trying to use Kiwi to install stuff for blog rolling with node js... Any one also get the 404 error I am getting on installing the seeds like haml ?
[01:41] itissid: Are there services down?
[01:42] stagas: itissid: kiwi is no longer developed, use npm
[01:43] itissid: Damn...
[01:43] itissid: Is wheat the best blogging engine on nodejs?
[01:43] stagas: itissid: http://github.com/isaacs/npm
[01:43] itissid: Any ecommendations?
[01:44] melpad: Roll one yourself
[01:51] eisd has joined the channel
[01:51] jesusabdullah: itissid: wheat's the only one I'm aware of, and afaik it ain't bad
[01:56] eisd_ has joined the channel
[01:58] itissid: Does any one know this error on NPM ucs {bad_utf8_character_code}: npm
[01:58] itissid: ?
[01:59] itissid: On NPM installation
[01:59] abiraja has joined the channel
[02:04] bentruyman has joined the channel
[02:08] mjr_ has joined the channel
[02:10] jedschmidt has joined the channel
[02:11] jedschmidt: isaacs: i'm getting a "Error: ucs {bad_utf8_character_code}: socket.io" on `npm install socket.io`... any advice you can give me?
[02:12] itissid: jedschmidt: I am getting same error on npm installation..
[02:12] nroot7: I am getting utf8 encoding error while installing npm ?
[02:12] jedschmidt: itissid: ah, good point.
[02:13] itissid: Ok so 4 people 
[02:13] itissid: 3*
[02:17] itissid: Any one?
[02:18] huyhong has joined the channel
[02:19] mikelsi has joined the channel
[02:20] jshirley has joined the channel
[02:22] paul__ has joined the channel
[02:22] benburkert has joined the channel
[02:22] meso has joined the channel
[02:22] paul__ has joined the channel
[02:24] murphy has joined the channel
[02:25] manveru: same error here
[02:25] saikat has joined the channel
[02:26] charly1 has joined the channel
[02:27] itissid: Still No one! Oh com'on its 4 of us now!! :)
[02:27] alexb_ has joined the channel
[02:29] stagas: itissid: you can still grab modules from github and use them without npm
[02:30] itissid: stagas: Ofcourse... But that does not solve this problem...
[02:30] manveru: i bet 9a0adec442b7a09a9a49604307fb09331140c7c4 broke it
[02:31] bruse: use git bisect
[02:31] bruse: and then git blame!
[02:31] manveru: there's an issue associated
[02:32] manveru: http://github.com/isaacs/npm/issues/issue/258
[02:33] samullen has joined the channel
[02:33] samullen: who
[02:34] noahcampbell has joined the channel
[02:36] dipser has joined the channel
[02:39] itissid: the header put in a request to : registry.npmjs.org/
[02:39] itissid: IN the file request.js
[02:39] itissid: where the error is
[02:39] itissid: the browser gives {"error":"ucs","reason":"{bad_utf8_character_code}"}
[02:40] itissid: on that address...
[02:40] mr_daniel has joined the channel
[02:40] probablycorey has joined the channel
[02:41] hzin has joined the channel
[02:42] huyhong1 has joined the channel
[02:42] samullen: Hey, I have a question about developing for Node. Mainly about environment.
[02:42] samullen: Is there a better way than just coding something up and restarting the server?
[02:42] samullen: What do most people use?
[02:43] manveru: itissid: ok, then my bet is wrong :)
[02:44] itissid: isaacs :Is there something wrong with this  registry.npmjs.org service? 
[02:45] deepthawtz has joined the channel
[02:47] samuel has joined the channel
[02:49] hzin has joined the channel
[02:49] c4milo has joined the channel
[02:49] c4milo1 has joined the channel
[02:49] samuel: I'd like to know what people are using to develop in node. 
[02:49] samuel: Is there an IRB-like environment to test things in?
[02:49] samuel: Can I just write a script?
[02:50] Tim_Smart: samuel: `node` should open the repl
[02:50] tekky has joined the channel
[02:50] hzin has joined the channel
[02:51] samuel: Tim_Smart: thanks, maybe I should just ask for a decent getting started guide.
[02:51] samuel: I have an idea that I originally wrote in mod_perl, but it would be better in node.
[02:52] samuel: I just really don't know where to get started.
[02:53] jesusabdullah: What idea? ooc?
[02:54] unomi has joined the channel
[02:54] samuel: jesusabdullah: nothing big, just something to help out with JS/CSS minification. Yes, I know there are projects which do that.
[02:58] samuel has left the channel
[03:00] derferman has joined the channel
[03:00] dipser_ has joined the channel
[03:00] itissid: isaacs :Is there something wrong with this  registry.npmjs.org service? 
[03:04] Tim_Smart: registry.npmjs.org got accidentally "{"error":"ucs","reason":"{bad_utf8_character_code}"}"
[03:09] derencius has joined the channel
[03:11] tmpvar: peace
[03:12] bentruyman has joined the channel
[03:13] unomi: does Express try to do something fancy with outgoing html <tags><like/><these/></tags> ?
[03:13] unomi: it seems as though if there isn't a space between the tags, then it tries to make them all pretty
[03:13] itissid: Tim_Smart: Yeah that is down...
[03:14] itissid: apparently
[03:14] itissid: Any one know how to install wheat without npm?
[03:17] badaxx_ has joined the channel
[03:17] unomi: just want to confirm that this is an express artifact and not something else
[03:19] cataska has joined the channel
[03:20] abiraja has joined the channel
[03:20] steadicat has joined the channel
[03:22] sonnym has joined the channel
[03:23] Me1000 has joined the channel
[03:23] elliottcable: JimBastard_ 3↼ what?
[03:23] elliottcable: 04:50:20 #Node.js: < JimBastard_> elliottcable: you won alaska
[03:25] SubStack: elliottcable: I got 2nd place!
[03:26] huyhong has joined the channel
[03:29] jamescarr_: hmmm
[03:29] jamescarr_: need a name for a web developer meetup group ;)
[03:30] aho: james carr and the funky bunch
[03:30] aho: np
[03:30] aho: <:
[03:31] tmm1: JimBastard_: nice blog post
[03:36] zpoley has joined the channel
[03:37] jakehow has joined the channel
[03:38] jamesarosen has joined the channel
[03:41] langworthy has joined the channel
[03:43] aconbere has joined the channel
[03:47] isaacs has joined the channel
[03:48] isaacs: npm registry seems to be down
[03:49] itissid: isaacs: so how much time to get it up?
[03:50] isaacs: itissid: not sure. 
[03:50] isaacs: some kind of couchdb errro
[03:50] whoahbot has joined the channel
[03:50] itissid: YEah we were discussing that ^
[03:51] itissid: There is a thread on that some where... http://mail-archives.apache.org/mod_mbox/couchdb-dev/200905.mbox/%3C739883689.1241648730276.JavaMail.jira@brutus%3E
[03:51] itissid: is this related?
[03:53] whoahbot has joined the channel
[03:53] isaacs: interesting.
[03:54] CrabDude has joined the channel
[03:55] itissid: I hope you can track it down...
[03:55] derferman: what is the best websocket server for node right now?
[03:55] derferman: I am looking for simplicity
[03:56] SubStack: socket.io has the best fallbacks
[03:57] SubStack: then there's node-websocket-server
[03:57] itissid: derferman: Better go for something that is complete 
[03:57] isaacs: itissid: yeah... weird.
[04:00] Blink7 has joined the channel
[04:00] SubStack: derferman: also if you like awesome abstractions I wrote dnode which can use socket.io
[04:00] derferman: substack, looks like socket.io is going to be perfect for what I need
[04:01] jesusabdullah: dnode is pretty sweet, fyi
[04:01] jesusabdullah: I'm trying to port it to python right now
[04:01] jesusabdullah: No telling if it'll ever be usable, but
[04:01] jesusabdullah: It's an interesting experience!
[04:01] SubStack: and I'm porting to ruby
[04:01] isaacs: itissid: i can't seem to find any high-ascii chars in the output, though
[04:01] isaacs: the changes feed is still valid...
[04:02] SubStack: it's going to rock so much once you can glue together a bunch of services with it written in different languages
[04:03] shaver: SubStack: like with Thrift, I guess
[04:03] SubStack: I've heard that mentioned
[04:03] SubStack: oh right this
[04:03] itissid: isaacs: I dont know a lot about this ... but is there not a diff between the rendering on the terminal(your viewing console)..? Could it not show interpret ascii charachters correctly...
[04:03] SubStack: yeah, dnode is more... javascripty
[04:04] SubStack: less crufty java baggage
[04:04] shaver: is there a thrift binding for node?
[04:04] jesusabdullah: thrift?
[04:04] nsm has joined the channel
[04:04] Aria: <wince> Thrift.
[04:04] Aria: No, there's not that I'm aware of.
[04:04] jesusabdullah: What is Thrift?
[04:04] Aria: It's a serialization format
[04:04] jesusabdullah: ah
[04:04] Aria: Really kinda scary crufty though
[04:04] jesusabdullah: Sounds familiar
[04:04] jesusabdullah: Ah
[04:04] shaver: like protobufs, kinda, but with some RPC on it
[04:05] jesusabdullah: S:
[04:05] Aria: Yeah
[04:05] SubStack: anyways, you'll be able to serve up a service with dnode in any supported lang, and you've got a cross-platform api for free
[04:05] SubStack: s/platform/language/
[04:05] rex_fernando has joined the channel
[04:05] SubStack: with callbacks too!
[04:06] SubStack: hmm I am not good at articulating this
[04:06] luchenbill has joined the channel
[04:06] Aria: Sounds good though!
[04:07] SubStack: so you can write a server in node and another one in perl and then a ruby process connects to both and makes some async remote method calls
[04:08] probablycorey has joined the channel
[04:08] isaacs: itissid: apparently it's timing out, not sure why.  {gen_server,call,[couch_query_servers,{get_proc,<<"javascript">>}]}
[04:08] sanduz2 has joined the channel
[04:09] itissid: Hmm this is your server right? Cant you reboot the service then? 
[04:10] derferman has joined the channel
[04:10] isaacs: itissid: it is not my server.
[04:10] isaacs: itissid: it's couchone.com
[04:10] isaacs: gonna have to wait for tomorrow morning, i guess.
[04:10] isaacs: that sucks
[04:11] itissid: yeah.. 
[04:11] isaacs: but, ya know, free, beggars and choosers and all that
[04:11] itissid: Anyways thanks for looking at it.. Atleast now we wont bang our heads, wondering....
[04:11] isaacs: np
[04:11] isaacs: spread the word.  i'm aware.  it's being looked into
[04:11] jesusabdullah: npm registry is down? bummer
[04:11] isaacs: take a day off :)
[04:12] itissid: :)
[04:12] isaacs: go outside, watch a movie, go to bed at a decent hour, etc.
[04:12] isaacs: :D
[04:12] jesusabdullah: hehe
[04:12] deepthawtz has joined the channel
[04:13] secoif: I'd like to connect to a legacy SOAP only :(  web-service using node.js, anyone got any tips or know of any libs that might help?
[04:13] shaver: boy
[04:14] shaver: I have that on my list too
[04:14] jesusabdullah: afaik, not soap libs yet
[04:15] blackdog_ has joined the channel
[04:15] jesusabdullah: If you're decent at C++, you can try making bindings to a c++ SOAP lib
[04:15] secoif: jesusabdullah: ooo
[04:15] shaver: C++ soap libs are rarely as async as you want for node.js
[04:15] shaver: very very rarely
[04:16] jesusabdullah: This is probably true. However, you could also avoid having to write your own SOAP client that way
[04:16] jesusabdullah: which I bet is annoying
[04:16] shaver: and binding multiple async systems is a great way to end up dirty and drunk on the street corner, talking to yourself about the XS-* stack
[04:17] jesusabdullah: XD
[04:18] jesusabdullah: Google seems to suggest that using xmlhttprequest or some BS makes it "easy" to consume SOAP webservices
[04:18] shaver: (WS stack)
[04:18] jesusabdullah: They're totally full of shit, though
[04:18] shaver: sure, the same way that netcat makes it easy to consume SSL
[04:18] jesusabdullah: exactly XD
[04:20] jesusabdullah: http://stackoverflow.com/questions/124269/simplest-soap-example-using-javascript
[04:20] jesusabdullah: >_<
[04:22] rex_fernando has joined the channel
[04:22] saikat has joined the channel
[04:23] langworthy has joined the channel
[04:23] secoif: hrm, none of these solutions look simple at all. dang.
[04:25] jesusabdullah: Just like SOAP!
[04:26] secoif: hrm, since I'm running apache on the same server, perhaps I could pass the requests through a local php script built on this
[04:26] secoif: http://snipplr.com/view/30223/getful-soap-to-json/
[04:26] rex_fernando_ has joined the channel
[04:27] rex_fernando_ has left the channel
[04:27] rex_fernando_ has joined the channel
[04:29] shaver: "I know, I'll use SOAP" is one of those "now you have two problems" sorts of situations
[04:30] secoif: unfortunately it's an external system we have no control over
[04:31] NemesisD has joined the channel
[04:31] jesusabdullah: I would personally use an intermediate service, but I'm also lazy and never expect a lot of connections/load
[04:32] jesusabdullah: You might also be able to port that to pure js
[04:33] alexb_: Why is soap so evil for node, isn't it basically just an XML schema
[04:34] manveru: alexb_: not just for node, anywhere
[04:35] jesusabdullah: xml schemas suck
[04:36] jesusabdullah: typically
[04:36] jesusabdullah: because they only really give you a way to make "containers"
[04:36] jesusabdullah: <thing> like, this is in a "thing" container </thing>
[04:36] jesusabdullah: and you need a lot of those, along with a bunch of outside definition, to make it useful
[04:36] alexb_: If the wsdl and the soap library is there, i find it can be a bit faster than coding my own library
[04:37] alexb_: Not that i like the standard 
[04:37] jesusabdullah: and then there's that soap is known for having cruft
[04:37] alexb_: Soap extensions i understand because those are over complicated had not defined well
[04:39] aconbere has joined the channel
[04:42] emmanueloga has joined the channel
[04:52] keeran has joined the channel
[04:54] benburkert has joined the channel
[04:54] emmanueloga has joined the channel
[04:56] vmthehut has joined the channel
[04:59] rex_fernando has joined the channel
[05:07] Ezku_ has joined the channel
[05:09] mjr__ has joined the channel
[05:12] siculars: so whats up with http://registry.npmjs.org/ barfing 
[05:14] Aria: isaacs has been looking into it
[05:14] jesusabdullah: host is having promblems, isaac is aware
[05:15] siculars: according to the error looks like someone tried to publish something couch didnt like ... {"error":"ucs","reason":"{bad_utf8_character_code}"}
[05:16] mjr__: isaacs: do you have the couch hosting guy's email?
[05:18] a_meteorite has joined the channel
[05:21] derferman has joined the channel
[05:22] ysynopsis1 has joined the channel
[05:28] cataska has joined the channel
[05:31] hzin has joined the channel
[05:37] probablycorey has joined the channel
[05:38] probablycorey has joined the channel
[05:40] sudoer has joined the channel
[05:41] NemesisD has joined the channel
[05:42] unomi has joined the channel
[05:44] hassox has joined the channel
[05:49] jtoy has joined the channel
[05:50] huyhong1 has joined the channel
[05:51] jchris has joined the channel
[05:55] itissid: I am a trying to edit my HAML files with VIM for use with express... Are tehre any indentation quircks?
[05:57] chetan51 has joined the channel
[05:58] chetan51: Hi guys, so I've written a minimalistic Node.js framework inspired by Code Igniter. I believe that people will find it useful, if only they knew about it. How do I spread the word about it? What channels / forums / etc can I post to about it? Here it is at GitHub: http://github.com/chetan51/ni
[05:59] chetan51: Any advice will be appreciated!
[05:59] SubStack: chetan51: add it to http://github.com/ry/node/wikis/modules 
[06:00] chetan51: I've done that, but there are just so many that it's lost in the huge list
[06:02] SubStack: blog about it too
[06:03] SubStack: then there is http://www.reddit.com/r/node and the mailing list
[06:03] SubStack: hmmm where did announcer go...
[06:03] pquerna: add it to npm?
[06:03] SubStack: oh right that too
[06:04] SubStack: npm is the best
[06:05] SubStack: 
ACTION wishes npmjs.org was up already

[06:06] jesusabdullah: ni? Cute name.
[06:09] chetan51: Awesome, I'll do all those things
[06:09] chetan51: @jesusabdullah: Haha, thanks, I think so too.
[06:10] gerred has joined the channel
[06:16] path[l] has joined the channel
[06:18] stephenjudkins_ has joined the channel
[06:19] itissid: hey is the nodejs HAML implementation ok to use with express?
[06:22] jesusabdullah: chetan51: I actually like the look of this. Nice that it gives you structure without embedding you in some Way Of Doing Things
[06:22] jesusabdullah: *watch*
[06:22] ctp has joined the channel
[06:24] chetan51: @jesusabdullah: Thanks, I'm glad you like it! That was my vision. If you use it, please try to contribute and make it better! :)
[06:24] alexb_: If node is doing a large post request to a slow server, will the post block?
[06:31] jesusabdullah: chetan51: You got it :v
[06:33] shaver: alexb_: post handling looks stream-based, so no
[06:34] deepthawtz has joined the channel
[06:36] itissid: hey did someone have a problem with the same/old page being served by express even though the server was restarted after changing the HAML/HTML file...?
[06:36] itissid: Its like its reading from a cache thats not refreshed....
[06:37] evl has joined the channel
[06:37] itissid: Any ideas why this might be? How i can correct it
[06:37] nroot7 has joined the channel
[06:37] _CrabDude has joined the channel
[06:40] xla has joined the channel
[06:40] sugardave has joined the channel
[06:46] tobiassjosten has joined the channel
[06:46] hoodoos: hey guys, what is the way to correctly log Exception in node? sys.puts(JSON.stringify(ex)) won't work always, cuz exceptions might have circular dependencies from time to time..
[06:46] hassox has joined the channel
[06:47] whoahbot has joined the channel
[06:47] hoodoos: so the question actually is how to correctly serialize Exception object to obtain all possible information from it
[06:48] omarkj has joined the channel
[06:48] pquerna: hoodoos: there was a thread on nodejs-dev recently about this...
[06:49] Aria: console.log(exception) ?
[06:49] pquerna: sys.inspect
[06:49] jesusabdullah: Yeah, probably sys.inspect
[06:50] jesusabdullah: or, even
[06:50] jesusabdullah: console.log(sys.inspect(exception))
[06:50] Aria: console.log already inspects.
[06:50] whoahbot has joined the channel
[06:50] Aria: Though you'll want the stack, too: console.log(e, e.stack)
[06:51] whoahbot has joined the channel
[06:51] Aria: And on that note, I'm going to bed.
[06:52] Aria: Good night!
[06:52] whoahbot has joined the channel
[06:54] sveimac has joined the channel
[06:54] whoahbot_ has joined the channel
[06:57] martin-g has joined the channel
[07:01] hassox has joined the channel
[07:01] nroot7: how can someone create a local copy (or mirror) of npm
[07:02] hannesw_ has joined the channel
[07:02] ivanfi has joined the channel
[07:05] jesusabdullah: nroot7: You mean the registry?
[07:05] spetrea has joined the channel
[07:06] ph^ has joined the channel
[07:06] nroot7: I mean like right now npm is down.. so in conditions like this we can have a backup
[07:06] mif86 has joined the channel
[07:08] SubStack: hoodoos: would just .toString() do the trick?
[07:09] SubStack: unless you care about stack traces
[07:09] herbySk has joined the channel
[07:09] hoodoos: SubStack, sure I care :)
[07:10] SubStack: well
[07:10] SubStack: e.message && e.stack ? e.message + '\n' + e.stack : e.toString())
[07:10] SubStack: might be good enough
[07:10] SubStack: I've done stuff like this elsewhere anyways
[07:10] jesusabdullah: I didn't think console.log inspected
[07:10] jesusabdullah: Hmm
[07:10] jesusabdullah: Maybe that's new
[07:11] SubStack: of course, stack and message /could/ contain circular refs too
[07:11] hoodoos: sometimes console log is innaccessible, maybe I'm wrong though..
[07:11] hoodoos: yea, that's the problem
[07:11] BryanWB has joined the channel
[07:11] SubStack: but you just get '[Circular]'
[07:11] SubStack: which isn't so bad
[07:11] hoodoos: on sys.inspect?
[07:11] dipser has joined the channel
[07:11] iheartnodejs has joined the channel
[07:11] SubStack: just on console.log
[07:11] iheartnodejs: hoo!
[07:12] hoodoos: okay, i'll try console log out
[07:12] SubStack: sys.inspect too
[07:12] SubStack: v8: var c = {}; c.c = c; sys.inspect(c)
[07:12] v8bot: SubStack: ReferenceError: sys is not defined
[07:12] delapouite has joined the channel
[07:12] SubStack: -_-
[07:12] SubStack: v8: var c = {}; c.c = c; console.log(c)
[07:12] v8bot: SubStack: ReferenceError: console is not defined
[07:12] SubStack: useless!
[07:12] jetienne has joined the channel
[07:12] SubStack: v8: var sys = require('sys'); var c = {}; c.c = c; sys.inspect(c)
[07:12] v8bot: SubStack: ReferenceError: require is not defined
[07:13] abiraja has joined the channel
[07:13] hzin has joined the channel
[07:14] herbySk: v8: Object.keys(({}).valueOf.call())
[07:14] v8bot: herbySk: []
[07:14] herbySk: (seems there are no keys in global :-| )
[07:15] hassox has joined the channel
[07:18] hdon has joined the channel
[07:18] SamuraiJack has joined the channel
[07:19] abiraja has joined the channel
[07:19] benreesman has joined the channel
[07:21] derferman has joined the channel
[07:22] boboroshi has joined the channel
[07:22] tilgovi has joined the channel
[07:23] shockie has joined the channel
[07:24] whoahbot_ has joined the channel
[07:25] admc has joined the channel
[07:26] whoahbot has joined the channel
[07:27] whoahbot has joined the channel
[07:28] whoahbot has joined the channel
[07:28] TomY_ has joined the channel
[07:29] emmanueloga has joined the channel
[07:29] whoahbot has joined the channel
[07:31] itissid has joined the channel
[07:33] BryanWB has joined the channel
[07:34] abiraja has joined the channel
[07:35] royi has joined the channel
[07:37] ooooPsss has joined the channel
[07:37] BryanWB has joined the channel
[07:38] whoahbot has joined the channel
[07:39] adambeynon has joined the channel
[07:40] saikat has joined the channel
[07:43] boboroshi has joined the channel
[07:47] margle has joined the channel
[07:49] viktors has joined the channel
[07:49] nroot7: I just cloned howtonode.org. how do I run its app.js
[07:49] nroot7: when I run using node app.js it simply returns
[07:51] ntelford has joined the channel
[07:51] hdon has joined the channel
[07:53] royi has left the channel
[07:54] dgathright has joined the channel
[07:55] javajunky has joined the channel
[07:57] derferman has joined the channel
[07:58] matjas has joined the channel
[08:00] rnewson has joined the channel
[08:00] rnewson has joined the channel
[08:01] mlangenberg has joined the channel
[08:04] boboroshi has joined the channel
[08:04] aubergine has joined the channel
[08:05] mlangenberg: Hi all
[08:09] jetienne_ has joined the channel
[08:11] aubergine has joined the channel
[08:12] aabt has joined the channel
[08:14] aubergine_ has joined the channel
[08:16] rauchg_ has joined the channel
[08:22] emmanueloga has joined the channel
[08:22] kloeri has joined the channel
[08:22] dgathright_ has joined the channel
[08:22] ajpiano has joined the channel
[08:27] mlangenberg: hmm... not so much activity in this channel. I wished I could talk to someone who has some experience with Socket.IO and/or Faye.
[08:27] SubStack: specific questions work best
[08:28] mlangenberg: Well it's kinda architectural. I have a Rails app, for which I would like to add chat functionality. 
[08:29] mlangenberg: There are n organizations each having n chat boxes with clients.
[08:29] SubStack: if you wait a few days I'll have dnode-ruby done by then
[08:29] SubStack: then you can use dnode's socket.io adapter to proxy directly into ruby
[08:30] mlangenberg: So in that case, Socket.IO would be my best bet to base my chat backend on.
[08:30] caolanm has joined the channel
[08:31] mlangenberg: By the way, do you perhaps know which has the best browser client support? Socket.IO or Faye?
[08:31] SubStack: I haven't used faye but I do know that socket.io works in IE5.5
[08:32] SubStack: which is insanity
[08:32] bpot has joined the channel
[08:32] SubStack: well at least it used to before that JSON adapater got moved or removed
[08:34] martin-g has left the channel
[08:34] sideshowcoder has joined the channel
[08:35] ker2x has joined the channel
[08:35] ker2x: hi :)
[08:35] mlangenberg: hello
[08:35] ker2x: i ahve a probleme with npm, it always worked correctly afaik
[08:36] ker2x: and now : message: 'ucs {bad_utf8_character_code}:
[08:36] ker2x: any tought ?
[08:37] stephank has joined the channel
[08:37] Tim_Smart has joined the channel
[08:37] _CrabDude: anyone know how to do basic authentication with http.createClient? (e.g., http://user:pw@abc.com)
[08:37] ker2x: http://npm.mape.me/  do no list any package too
[08:38] mape: ker2x: the repo is down
[08:38] ker2x: the repository is down or something ?
[08:38] ker2x: okay thx
[08:38] Kami__: _CrabDude, yes, you need to send autherization header with base64 encoded username and password
[08:38] ker2x: that explain :)
[08:39] ker2x: mape: need help to create mirrors or something ? considering we're using nodejs in production, i could ask my boss if we could dedicate some server ressource
[08:39] sideshowcoder has joined the channel
[08:39] mape: ker2x: you should talk to isaacs 
[08:39] _CrabDude: Kami__ ahhh, yes, thx. I was getting caught up on the credentials param in createClient. thx again
[08:39] ker2x: ok thx
[08:40] Kami__: _CrabDude, np, credentials parameters in createClient are for ssl stuff
[08:40] overra has joined the channel
[08:41] martin-g has joined the channel
[08:45] robotarmy has joined the channel
[08:46] markwubben has joined the channel
[08:48] satori_ has joined the channel
[08:48] evl has left the channel
[08:48] sideshowcoder has joined the channel
[08:48] aubergine_ has joined the channel
[08:51] aubergine has joined the channel
[08:56] ker2x: 
ACTION poked isaacs 

[08:57] ay has left the channel
[09:03] boboroshi has joined the channel
[09:03] q_no has joined the channel
[09:06] admc has joined the channel
[09:10] xla has joined the channel
[09:11] liesen has joined the channel
[09:12] derferman has joined the channel
[09:14] boboroshi has joined the channel
[09:16] kschzt has joined the channel
[09:16] kschzt: hey, is isaacs asleep? is npm registry broken for someone else?
[09:17] ker2x: yep, npm is broken
[09:19] kschzt: and I suppose only isaacs can fix it :) w00t
[09:21] badaxx has joined the channel
[09:40] ker2x: look like npm is back
[09:40] ker2x: isn't it ?
[09:40] ker2x: oops, nope
[09:42] industrial has left the channel
[09:43] adambeynon has joined the channel
[09:45] ker2x: npmjs support : We've raised $50.00, and we still need your support!
[09:45] ker2x: $50 :(
[09:52] rnewson has joined the channel
[09:59] ctp has joined the channel
[10:01] agnat has joined the channel
[10:06] mape: it isn't a money issue
[10:06] mape: although donations are always nice
[10:08] kschzt: some sort of utf8 thing, looks like the message comes from couch, maybe
[10:09] overra: woot doubled that $50
[10:10] viirya has joined the channel
[10:13] kschzt: phew, gladly i could copy my pkg setup from another environment 
[10:15] BryanWB has joined the channel
[10:19] path[l] has joined the channel
[10:19] liesen has joined the channel
[10:21] Kami__: if you need a source tarball for some npm module, nodul.es mirrors tarballs for latest version of each module
[10:22] omarkj has joined the channel
[10:31] aubergine_ has joined the channel
[10:33] aubergine_ has joined the channel
[10:33] robotarmy has joined the channel
[10:34] markwubben: isaacs: i'm finding package.json's without version numbers, wondering if npm should support github-based packages with versioning via github tags?
[10:36] aliem has joined the channel
[10:41] guitt has joined the channel
[10:45] jetienne_: where is stagas when i need him :)
[11:00] ekidd has joined the channel
[11:01] agnat has joined the channel
[11:02] virtuo has joined the channel
[11:04] omarkj has joined the channel
[11:04] mikew3c has joined the channel
[11:09] christophsturm has joined the channel
[11:10] sveimac has joined the channel
[11:11] fermion has joined the channel
[11:11] mAritz has joined the channel
[11:19] trochala has joined the channel
[11:19] trochala has joined the channel
[11:27] path[l] has joined the channel
[11:27] mikew3c has joined the channel
[11:32] jetienne_: q. anybody got url for a doc on socket io ? im trying to integrate it for 3h
[11:39] Neil has joined the channel
[11:45] q_no: jetienne_, the chat expample covers most of it... at least for me it was a great point to start learning from
[11:46] jetienne_: q_no: lucky you :) "Invalid url for WebSocket ws://localhost:8080/socket.io/websocket" means anything to you ?
[11:47] abiraja has joined the channel
[11:48] christophsturm: npm install expresso is giving me an error: https://gist.github.com/59ef76db3708758ad035
[11:49] Neil_ has joined the channel
[11:56] dpritchett has joined the channel
[11:59] agnat has joined the channel
[12:00] emmanueloga has joined the channel
[12:00] aubergine_ has joined the channel
[12:02] aubergine has joined the channel
[12:02] mlangenberg has joined the channel
[12:03] mape: christophsturm: the npm repo is down
[12:03] c4milo has joined the channel
[12:05] LFabien has joined the channel
[12:07] xla has joined the channel
[12:17] MattJ has joined the channel
[12:28] devkorcvince has joined the channel
[12:31] xla has joined the channel
[12:35] liesen has joined the channel
[12:35] shimondoodkin has joined the channel
[12:36] shimondoodkin: hello, does anybody knows of a good JSON.stringify function with citcular references
[12:40] dpritchett has joined the channel
[12:42] dpritchett: Does anyone have a good example of DNode in action and why it's so darned useful to pass functions back and forth between server and client?
[12:44] mlangenberg: dpritchett: smells like Java's RMI to me *yuk*
[12:45] dpritchett: Yes, it was directly linked to RPC/RMI.  I'm reading the wikipedia article on RPC but so far I'm not having a lightbulb moment.
[12:49] dpritchett: I guess it's useful when you want to perform a calculation or access a resource that the remote machine can get at better than the local one
[12:52] dpritchett: ah, now i get it
[12:52] dpritchett: Facebook's Dinosaur Island and it's Thrift server use something like RPC to communicate between the game server and the various AI bots playing the game
[12:52] dpritchett: http://www.facebook.com/careers/puzzles.php?puzzle_id=19
[12:53] bruse: RPC is a pretty broad term
[12:53] dpritchett: I understood it in principle I was just suffering from a lack of creativity in figuring out why it would be so useful
[12:54] matjas has joined the channel
[12:54] dpritchett: i guess it's basically about programs that interface directly with one another via sockets rather than using a messaging system or a common data store
[12:56] bruse: when i wrote a system to parse websites i split it into parts where each parser for every website was its own separate process that used XML-RPC to talk to a server which then crunched the extracted data and inserted what i wanted into the database
[12:57] bruse: that way i could write the language in any language i wanted
[12:57] bruse: and also distribute the parsers to other servers if necessery
[13:01] Anti-X has joined the channel
[13:05] rcy has joined the channel
[13:08] jtsnow has joined the channel
[13:08] stephank: Ooh, llvm 2.8 is calling its C++ support production ready.
[13:08] phiggins has joined the channel
[13:09] micheil has joined the channel
[13:10] dpritchett: thanks bruse
[13:10] Akufen has joined the channel
[13:11] paul___ has joined the channel
[13:13] wink_ has joined the channel
[13:18] aubergine_ has joined the channel
[13:18] rauchg_ has joined the channel
[13:19] adambeynon has joined the channel
[13:20] jherdman has joined the channel
[13:20] aubergine_ has joined the channel
[13:20] evanpro has joined the channel
[13:21] rauchg_ has joined the channel
[13:23] Nikita_ has joined the channel
[13:24] hzin has joined the channel
[13:25] sonnym has joined the channel
[13:26] dnolen has joined the channel
[13:27] peritus- has joined the channel
[13:27] karboh has joined the channel
[13:28] peritus-: Hey. I'm getting 'node(28033) malloc: *** error for object 0x1008257f8: incorrect checksum for freed object - object was probably modified after being freed' when using crypto.createHmac on mac os x .. how do I debug such a thing?
[13:28] wink_: pertius: run it through valgrind and use node_g
[13:29] wink_: it'll run slower, but it should show you what loc allocated the memory
[13:30] peritus-: wink_: sure. thanks. i'll do that.
[13:31] wink_: once you get the output, pastebin it and i'll help where i can
[13:34] peritus-: wink_: i'm low on battery and thus can't build valgrind :( .. but i'll do as soon as I get near my charger.
[13:38] peritus-: wink_: googling 'node_g' doesn't raise proper results .. where do I read up on that one ?
[13:41] davidsklar has joined the channel
[13:41] ben_alman has joined the channel
[13:41] ntelford has joined the channel
[13:42] ben_alman has joined the channel
[13:43] ben_alman has joined the channel
[13:43] ben_alman has joined the channel
[13:43] EyePulp has joined the channel
[13:44] ben_alman has joined the channel
[13:45] ben_alman has joined the channel
[13:45] ceej has joined the channel
[13:45] ben_alman has joined the channel
[13:46] figital has joined the channel
[13:46] agnat has joined the channel
[13:46] ben_alman has joined the channel
[13:46] mlangenberg: So funny how there is a movement from XMLRPC -> REST XML -> REST JSON -> JSON-RPC
[13:47] ben_alman has joined the channel
[13:47] ben_alman has joined the channel
[13:48] ben_alman has joined the channel
[13:49] ben_alman has joined the channel
[13:49] ben_alman has joined the channel
[13:50] ben_alman has joined the channel
[13:50] micheil has joined the channel
[13:50] ben_alman has joined the channel
[13:51] ben_alman has joined the channel
[13:52] ben_alman has joined the channel
[13:52] hannesw_ has joined the channel
[13:52] ben_alman has joined the channel
[13:53] fabeat has joined the channel
[13:53] ben_alman has joined the channel
[13:54] ben_alman has joined the channel
[13:54] ben_alman has joined the channel
[13:55] codelotus has joined the channel
[13:56] ben_alman has joined the channel
[13:56] ben_alman has joined the channel
[13:57] c4milo has left the channel
[13:57] peritus__ has joined the channel
[13:58] ben_alman has joined the channel
[13:59] jlilly has joined the channel
[13:59] 5EXABJ94L has joined the channel
[13:59] jansc has joined the channel
[14:00] nerdEd has joined the channel
[14:00] 5EXABJ94L has left the channel
[14:01] justinlilly has joined the channel
[14:02] justinlilly has joined the channel
[14:02] gerred has joined the channel
[14:06] c4milo has joined the channel
[14:08] aho has joined the channel
[14:08] aheckmann has joined the channel
[14:11] nefD has joined the channel
[14:12] davidwalsh has joined the channel
[14:12] probablycorey has joined the channel
[14:12] aconbere has joined the channel
[14:13] matt_c has joined the channel
[14:13] cefn has joined the channel
[14:16] bradleymeck has joined the channel
[14:17] micheil: anyone here played with riak?
[14:19] boaz has joined the channel
[14:19] dylang has joined the channel
[14:20] tg has joined the channel
[14:23] JimBastard_: thanks tmm1 , you got a top ranking for sure
[14:23] JimBastard_: :p
[14:27] V1 has joined the channel
[14:27] roger_raymond has joined the channel
[14:28] cefn: Only been on 10 mins, but no discussions on issue #266 and nothing in the room topic. Is anything likely to be done to bring the repository back soon? http://github.com/isaacs/npm/issues#sort=votes.
[14:30] kschzt: cefn, we have to wait until isaacs wakes up
[14:31] kschzt: it's 7.30 in SF so I'm thinking it'll take another 2h before it gets fixed
[14:31] cefn: Thanks :kschzt and noted nodul.es workaround in the logs. 
[14:32] sveimac has joined the channel
[14:32] Anti-X: 
ACTION slaps sveimac around a bit with a large trout

[14:33] micheil: kschzt, cefn: I've also tried to get in contact with him, but yeah, another 2h probably
[14:33] Anti-X: how unprofessional of him to sleep and stuff
[14:34] joncurra: sleep is for the weak!
[14:34] Anti-X: you can sleep on weekends
[14:34] joncurra: not if you're on irc
[14:35] joncurra: you're expected to be up
[14:37] Anti-X: i should make an irc bot that reads your mind and can answer for you while you sleep
[14:37] bradleymeck: 
ACTION flops about

[14:37] brainproxy has joined the channel
[14:38] dylang: this is my first to learn that isaacs sleeps.
[14:38] micheil: Anti-X: then would you have to have a night cap?
[14:38] trotter has joined the channel
[14:38] Anti-X: or you could just do the matrix thing
[14:38] figital has joined the channel
[14:38] Anti-X: with a spike in the back of your head
[14:38] Anti-X: but yeah i see the double pun
[14:40] JimBastard_: reddit has turned into such a shit show
[14:40] JimBastard_: http://www.reddit.com/r/programming/comments/dmbhm/calculating_the_social_influence_of_github/
[14:40] mif86 has joined the channel
[14:40] dylang: has anybody seen a node service to make sure a server is up and working?  could be websocket based.
[14:40] JimBastard_: dylang: yeah
[14:41] JimBastard_: pulse by Atmoz 
[14:41] micheil: dylang: there was one in node knockout
[14:41] JimBastard_: http://github.com/atmos/http-pulse-app
[14:42] dylang: micheil: JimBastard_: oh yeah, thanks for reminding me.
[14:42] JimBastard_: i dont think it would be hard to build one though
[14:42] JimBastard_: setInterval + http.Client
[14:42] c4milo: TJ ? 
[14:42] JimBastard_: santa?
[14:43] Yuffster has joined the channel
[14:43] dylang: JimBastard_: i was thinking of using websockets to avoid creating new connections, and possibly having a way to get server status even if something crazy happens like too many connections.
[14:44] JimBastard_: dylang: well then your server needs to have websocket server
[14:44] ehaas has joined the channel
[14:44] liesen has joined the channel
[14:45] dylang: JimBastard_: yup.  i have a couple ideas that will require a websocket server.  you think people would use a module that spawns a socket server or does it feel like overkill?
[14:46] JimBastard_: i dont see the benefit of using websockets versus polling for http status
[14:46] peritus- has joined the channel
[14:46] JimBastard_: i wouldnt think having a ping http request every 3 seconds would really amtter
[14:46] JimBastard_: and without websockets, you can poll sites that have no existing infrastrcture
[14:47] softdrink has joined the channel
[14:49] aubergine_ has joined the channel
[14:49] micheil: websockets are seriously over hyped.
[14:49] jansc has joined the channel
[14:49] micheil: dylang: if you need a websocket server, I've got one.
[14:50] dylang: true.  but websockets are new, cutting edge, and not standard yet.  that's been my justificaiton for everything else i do in node, anyone else?  :)
[14:50] micheil: dylang: hmm, well, really look at why you need to use the tech, not why you want to use the tech
[14:51] kgf: +1
[14:52] tmpvar has joined the channel
[14:54] dylang: micheil: without crazy experimentation we risk limitting ourselves to only what we think we need a technology for.   dumb luck and foolish experimationation sometimes can lead to the coolest inventions.
[14:54] cartab has joined the channel
[14:54] micheil: dylang: true, but for something like that I'm not sure
[14:54] micheil: dylang: if you write it similar to serverdensity it could be really cool
[14:55] micheil: in the sense that you start a node process running on one server, it opens a connection to another server and ping-pongs between the two ever X milliseconds
[14:55] micheil: say, every 500 milliseconds
[14:55] micheil: then you could have some data worth of push via websockets
[14:56] micheil: like if you used time-stamps as messages for ping/pong, then your server-app could calculate network latency and stuff
[14:56] micheil: although, you'd probably want to talk over UDP instead of TCP
[14:57] jpld has joined the channel
[14:58] zpoley has left the channel
[14:58] voodootikigod has joined the channel
[14:59] dylang: a cool idea.  i was thinking something like top that would focus on node, caching, data source connections for live servers.  i know there are tools for this but maybe not specificly for high-trafficed node-based projects that require zero latency.
[14:59] JimBastard_: 0 latency you say
[14:59] dylang: well
[14:59] JimBastard_: id like to purchase some of your wires 
[15:00] JimBastard_: i'll pay 5 dollars for one
[15:00] micheil: 0 latency isn't happening.
[15:00] JimBastard_: ohm agrees
[15:00] dylang: semantics - replace zero with low
[15:00] JimBastard_: replace websockets with polling
[15:00] JimBastard_: :-D
[15:00] dylang: replace node with ruby, php, or java.  and we're back to where we started.
[15:01] JimBastard_: it was a fun trip
[15:01] dylang: haha
[15:01] micheil: but no, definitely something which was a constant ping-pong could well be cool.
[15:01] nefD: so long, and thanks for all the fish
[15:01] JimBastard_: micheil: didnt i already post that pulse app? isnt that what that does?
[15:02] micheil: JimBastard_: yes, but you think I actually looked at it?
[15:02] micheil: :P
[15:02] Pilum has joined the channel
[15:02] Pilum has left the channel
[15:02] luchenbill has joined the channel
[15:03] omarkj: I found use for websockets, yay!
[15:03] peritus-: wink_: found a charger!
[15:03] Faustie has joined the channel
[15:03] micheil: hmm.. because the DOM is a tree, why should it matter that the HEAD comes before the BODY?
[15:03] peritus-: wink_: valgrind log is uploading as we speak
[15:03] micheil: omarkj: oh really?
[15:03] omarkj: micheil: Yup, using it to publish exchange market data.
[15:04] micheil: *shudders slightly*
[15:04] micheil: although, yeah, probably a good use.
[15:04] dipser: micheil: because this gets to load first
[15:04] omarkj: Haha.
[15:04] Yuffster has joined the channel
[15:04] micheil: dipser: ?
[15:05] micheil: oh, right.
[15:05] jansc has left the channel
[15:05] dshaw has joined the channel
[15:06] micheil: hey daniel.
[15:06] peritus-: wink_: ok, gist.github.com throws 413 Request Entity Too Large on 1.8mb logfiles
[15:06] micheil: peritus-: man... I wish I was that big. :P
[15:06] omarkj: I guess it's not the coolest use..
[15:06] micheil: I joke, I joke.
[15:06] aurynn has joined the channel
[15:06] dipser: this is why you should write tfoot right direct thead and before tbody of tables (html4)
[15:06] steadicat has joined the channel
[15:07] micheil: dshaw: was it you with the nvm HEAD issue?
[15:07] peritus-: wink_: ok, http://dl.dropbox.com/u/8735936/valgrind.log 
[15:08] Me1000 has joined the channel
[15:09] ows has joined the channel
[15:09] EyePulp has joined the channel
[15:11] ows has left the channel
[15:11] xla has joined the channel
[15:12] adambeynon has joined the channel
[15:13] aheckmann has joined the channel
[15:13] ows has joined the channel
[15:15] kriszyp has joined the channel
[15:17] scnd has joined the channel
[15:17] q_no has joined the channel
[15:22] EyePulp has joined the channel
[15:24] rick has joined the channel
[15:27] ajpiano has joined the channel
[15:29] ironfroggy_: I have built Node a few times but on a new machine I'm getting this when I try to configure: http://paste.pocoo.org/show/270938/
[15:29] LowValueTarget has joined the channel
[15:30] ironfroggy_: I have checked and python is built with threading and i can spawn threads myself, so i'm not sure what's going on
[15:30] isaacs has joined the channel
[15:30] loincloth has joined the channel
[15:32] ivanfi has left the channel
[15:34] dgathright has joined the channel
[15:35] gf3 has joined the channel
[15:36] spetrea has joined the channel
[15:38] tj has joined the channel
[15:38] jstemmer has joined the channel
[15:41] cefn has left the channel
[15:41] dgathright has joined the channel
[15:41] ajpiano has joined the channel
[15:43] jakehow has joined the channel
[15:43] ginader has joined the channel
[15:44] creationix has joined the channel
[15:45] agnat: ironfroggy_: what os is this? It fails to detect almost all libev related functions. what does config.log say?
[15:46] devkorcvince has joined the channel
[15:46] deepthawtz has joined the channel
[15:47] devkorcvince: Hello I need help my npm is acting weird I cant list all available package and install any of sassjs and hamljs
[15:47] isaacs: devkorcvince: registry is down.  looking into it.
[15:47] jesusabdullah: registry is acting up
[15:47] jesusabdullah: whups
[15:47] isaacs: :)
[15:47] jesusabdullah: isaacs BEAT ME
[15:47] isaacs: 
ACTION beats jesusabdullah 

[15:47] JimBastard_: isaacs: this is fucked up man, i want a refund for the month
[15:47] JimBastard_: i pay good money for access to npm
[15:47] devkorcvince: isaacs oh thanks
[15:48] aconbere has joined the channel
[15:48] JimBastard_: :p
[15:48] micheil: 
ACTION probably beat you all to it. gotta love twitter's dm's.

[15:48] rauchg_ has joined the channel
[15:48] micheil: rauchg_: want to start the day with an interesting bug?
[15:48] rauchg_: micheil: let's hear it
[15:49] micheil: okay, for some reason odd things are happening over flash sockets with socket.io
[15:49] bradleymeck: like?
[15:50] rauchg_: micheil: 
[15:50] jesusabdullah: I want to be COMPENSATED for DAMAGES over my LOST PRODUCTIVITY from not being able to use npm!
[15:50] jesusabdullah: stompy stompy
[15:50] rauchg_: yeah 
[15:50] micheil: rauchg_: in here?
[15:50] rauchg_: there's been issues with flashsocket i was gonna look into today
[15:50] ironfroggy_: agnat: its debian 5.0.4
[15:50] micheil: rauchg_: I'm thinking this might be fixable with the reworking of the websockets section.
[15:51] rauchg_: yup
[15:51] rauchg_: it's likely
[15:51] rauchg_: but it's also likely that it's shitty flash websocket
[15:51] micheil: true
[15:51] d0k has joined the channel
[15:51] micheil: that said, I found out that node-websocket-server has a pretty big issue in it..
[15:51] agnat: ironfroggy_: I'd suspect your missing important system headers ... it should find inotify, epoll, poll, select, eventfd anf probably a few others ... 
[15:51] micheil: I haven't been calling setTimeout(0) and no nagle
[15:52] ironfroggy_: and config.log has a bunch of different tracebacks in it. im not sure what is important and posting the whole thing seems useless. is it?
[15:52] micheil: rauchg_: also, have a look at the development branch of node-websocket-server, I think I've pushed the latest, but it allows you to create new websocket connections with just the manager, no need for a server, just a valid request.
[15:52] ironfroggy_: agnat: not surprising. i got handed a box that was stripped so bare i have to think they've actively just rm'ed everything that didn't break apt and left barely anything at all.
[15:52] rauchg_: micheil: i aw
[15:53] rauchg_: saw
[15:53] rauchg_: amazing
[15:53] steadicat has joined the channel
[15:53] micheil: rauchg_: that said, I'm not sure how you're actually managing connections within socket.io
[15:53] agnat: ironfroggy_: na, paste the whole thing ... let's have a look
[15:53] agnat: ironfroggy_: off-channel of course
[15:53] creationix has left the channel
[15:54] agnat: ironfroggy_: I guess clean install is not an option?
[15:55] ironfroggy_: http://fpaste.org/VVZz/
[15:56] ironfroggy_: agnat: no i didnt have much choice in the host in this case, because it had to be in sweden and i dont speak swedish
[15:56] agnat: ironfroggy_: heh
[15:57] bruse: i speak swedish fluently
[15:57] bruse: it's dead simple, honest
[15:57] ironfroggy_: i'm a stupid american. i feel stupid for being unilingual.
[15:57] bruse: i only know how to code in c
[15:57] bruse: i feel stupid for being unilingual too :(
[15:58] ironfroggy_: you are in #Node.js but you dont know javascript?
[15:58] bruse: i've been meaning to learn..
[15:59] bruse: actually, i know lots of languages, just not javascript that well
[15:59] kriskowal has joined the channel
[15:59] ironfroggy_: agnat: wish that log was formated better. too many long lines, sorry.
[15:59] wink_: bruse: its easy to learn enough to get in trouble, the intricacies are an infinitely deep hole however :>
[16:01] agnat: ironfroggy_: yeah, difficult to read ... anyway ... in fact you are missing all kinds of system headers and/or the system headers are just to old ... 
[16:01] bruse: it's amazing how much easier it is to get your point across in a language you barely know when you're drunk though
[16:01] bruse: or are we still talking about javascript?
[16:02] JimBastard_: jabascripts
[16:02] JimBastard_: tjholowaychuk knows
[16:02] jesusabdullah: That sounds like a good story
[16:02] jesusabdullah: "...and I don't even know c++!"
[16:02] JimBastard_: tjholowaychuk: you placed 2nd in "canada" btw
[16:02] ironfroggy_: agnat: ugh. ok. debian doesnt seem to have any easy to pick header package. i have to match kernel versions i guess..
[16:02] JimBastard_: after http://github.com/igrigorik
[16:03] agnat: ironfroggy_:  you could try to reinstall libc-dev and the kernel-libc headers ... but if the system is to old you'd have to hack node (or it's build system)
[16:03] loincloth has joined the channel
[16:03] gerred has joined the channel
[16:03] agnat: ironfroggy_: don't know the exact debian package names, though
[16:04] devkorcvince has joined the channel
[16:04] tjholowaychuk: JimBastard_: for what
[16:04] cardona507 has joined the channel
[16:04] tjholowaychuk: jabascript?
[16:04] tjholowaychuk: haha
[16:04] JimBastard_: tjholowaychuk: http://blog.nodejitsu.com/most-influential-github-users-by-location
[16:04] JimBastard_: its a little inaccurate, i talk about that in the blog
[16:04] JimBastard_: ill run it again
[16:05] JimBastard_: but you are top 5 in canada for sure
[16:05] caolanm: JimBastard_: what's it using an a measure of influence?
[16:05] JimBastard_: caolanm: the script is linked, its just doing followers / following now, i talk in the post about what can be done to make the results more accurate
[16:05] JimBastard_: i only spend a couple hours on this, got the idea from someone else
[16:05] JimBastard_: but i think its interesting / can be improved
[16:06] caolanm: JimBastard_: yeah, quite fun :)
[16:06] caolanm: a heatmap of github influence would be cool
[16:06] eazyigz has joined the channel
[16:06] tjholowaychuk: JimBastard_: haha crazy, interesting stuff
[16:06] loincloth has joined the channel
[16:06] agnat: ironfroggy_: I'd try to compile node in a fresh debian 5 vbox and copy over the binary ...
[16:06] eazyigz: what is the best way to spawn a 'true' unix daemon from nodejs?
[16:06] caolanm: though I think watchers is a better measure than followers
[16:06] JimBastard_: caolanm: yeah for sure
[16:07] tjholowaychuk: im probably the only person from victoria on there
[16:07] JimBastard_: also, project count
[16:07] JimBastard_: fork to own ratio
[16:07] JimBastard_: commits to other projects
[16:07] caolanm: tjholowaychuk: no. 1 in victoria!
[16:07] caolanm: 
ACTION wonders if he's no.1 in York

[16:07] JimBastard_: caolanm: you can run the script, ill run it quick if you want
[16:07] JimBastard_: lol
[16:08] JimBastard_: i dont really know ruby too well, or yahoo BossMan, so i wasnt able to make the modifications i wanted to (at least not in 10 minutes)
[16:08] micheil: tjholowaychuk: I'm probably the only one from australia still here :P
[16:08] JimBastard_: i think trying to geolocate the location field and then converting to long lat will be better
[16:08] caolanm: JimBastard_: I don't even know if I have ruby installed... probably do
[16:08] tjholowaychuk: micheil: haha :D. seems like canada is pretty weak on github
[16:09] JimBastard_: well, "York" isn't a good search query
[16:09] tjholowaychuk: repressenttt
[16:09] JimBastard_: will pick up on "New York"
[16:09] micheil: tjholowaychuk: :P
[16:09] caolanm: JimBastard_: d'oh
[16:09] caolanm: JimBastard_: york, uk  ?
[16:10] zpoley has joined the channel
[16:10] Sutto has joined the channel
[16:10] Sutto has left the channel
[16:11] ironfroggy_: agnat: not a bad idea. i'll give that a shot.
[16:11] JimBastard_: caolanm: ill try "York, UK"
[16:11] caolanm: JimBastard_: just keep narrowing your search until I'm at the top ;)
[16:11] JimBastard_: the problem with smaller regions is that people who follow a lot of projects tend to pollute them
[16:11] kdridi has joined the channel
[16:11] JimBastard_: caolanm: i was a nervous wreck the first time i calculated NY
[16:11] caolanm: haha
[16:12] JimBastard_: i kept saying to myself i would have to modify the algo if i didnt place, then i realized i could just tell people i placed anyway
[16:12] JimBastard_: no one would check
[16:12] JimBastard_: but then i did!
[16:12] JimBastard_: crisis averted!
[16:12] caolanm: \o/
[16:12] caolanm: hmm, has anyone actually checked that JimBastard_ placed in the top 5?
[16:12] JimBastard_: i gotta recalc NY thought with the normalization of location field 
[16:13] JimBastard_: caolanm: i had a mental image on the top 10 in ny based on who i know
[16:13] JimBastard_: and it was pretty accurate
[16:13] JimBastard_: i assumed i would place top 5
[16:13] JimBastard_: when i re-run it though i might get bumped to top 10
[16:13] JimBastard_: maybe not
[16:13] JimBastard_: i know jashkenas is top 5
[16:13] JimBastard_: and webiest should be removed
[16:14] JimBastard_: the script is very very basic
[16:14] spetrea has joined the channel
[16:14] spetrea: how do I melt 2 associative arrays ?
[16:16] phiggins: use php
[16:16] phiggins: js doesn't have associative arrays
[16:17] evl has joined the channel
[16:17] spetrea: phiggins: I'm sorry, wasn't looking for a flame war
[16:18] phiggins: i wasn't flaming...
[16:18] benburkert has joined the channel
[16:18] spetrea: phiggins: certainly switching the programming language is not the answer
[16:18] phiggins: do you want to mix two objects? (could be called dictionaries?)
[16:18] phiggins: or arrays
[16:18] phiggins: the answer is different 
[16:18] tjholowaychuk: spetrea: objects should be fine: { foo: 'bar', another: { foo: 'bar' } }
[16:19] rick has joined the channel
[16:19] rick has joined the channel
[16:19] tjholowaychuk: spetrea: Object.keys(obj); will give you ['foo', 'another'] so you can iterate
[16:20] spetrea: tjholowaychuk: yea two objects basically
[16:20] rick has joined the channel
[16:20] creationix has joined the channel
[16:20] phiggins: spetrea: https://gist.github.com/ab9bba733751c0643426
[16:20] rick has joined the channel
[16:20] rwaldron has joined the channel
[16:21] rwaldron has joined the channel
[16:23] mikeal has joined the channel
[16:23] sh1mmer has joined the channel
[16:23] jchris has joined the channel
[16:24] omarkj: Hmh. Is there no way to see form the response created client.request if the connection was rejected ?
[16:24] omarkj: (HTTP Client)
[16:24] felixge has joined the channel
[16:25] felixge: ryah: did you see my buffer leak test case?
[16:26] brianmario has joined the channel
[16:26] pquerna: omarkj: .on('error') ?
[16:26] omarkj: pquerna: Nope, that event isn't emitted.
[16:26] bradleymeck: and you are sure it is being rejected an not hung?
[16:27] bradleymeck: s/an/and/
[16:27] omarkj: bradleymeck: Yes. Got both services running here.
[16:28] omarkj: It's supposed to be refused by the way, but I need to catch it.
[16:28] bradleymeck: mmm (digs for a second)
[16:28] omarkj: Im digging as well. :)
[16:28] ph^ has joined the channel
[16:28] rnewson has joined the channel
[16:28] rnewson has joined the channel
[16:30] saimon_ has joined the channel
[16:31] tmpvar has joined the channel
[16:32] evanpro has joined the channel
[16:33] bartt has joined the channel
[16:33] mjr_ has joined the channel
[16:34] tmpvar has joined the channel
[16:34] bradleymeck: omarkj im getting ECONNREFUSED to fire, cant reproduce it not firing
[16:34] bradleymeck: anywho off to lunch
[16:34] felixge_ has joined the channel
[16:35] omarkj: bradleymeck: It's probably somewhere in my code, thanks for clarifying.
[16:35] hpoydar has joined the channel
[16:36] sudoer has joined the channel
[16:36] omarkj: Do you see where it's being fired?
[16:36] omarkj: got it
[16:38] siculars has joined the channel
[16:39] shockie has joined the channel
[16:40] JimBastard_: shockie: i love your internet video where you lip sing that song
[16:41] malkomalko has joined the channel
[16:42] V1 has joined the channel
[16:44] bpot has joined the channel
[16:45] JimBastard_: ohh hrmm oakland never made it to the post
[16:45] JimBastard_: mikeal and isaacs got on that one
[16:45] JimBastard_: #1: isaacs #2: byrnereese #3: jayallen #4: mikeal #5: substack 
[16:45] JimBastard_: SubStack too
[16:45] JimBastard_: SubStack: people love you in oakland
[16:45] mikeal: what up
[16:45] JimBastard_: mikeal: you are the #4 "most influential" github users in oakland
[16:46] mikeal: for real
[16:46] JimBastard_: yeah i just crunched oakland
[16:46] JimBastard_: i should update the blog post
[16:46] mikeal: i don't know who the other non-isaacs people are
[16:46] JimBastard_: its github user names
[16:47] tjholowaychuk: JimBastard_: what does the script use? just followers?
[16:47] isaacs: wow, i'm most influential in oakland?
[16:47] micheil: JimBastard: so, how were you working this stuff out?
[16:49] bartt has joined the channel
[16:49] JimBastard_: tjholowaychuk: yeah, the code is posted. i talk about the current flaws in the script too
[16:49] JimBastard_: i think it could be easily changed to be more accurate
[16:49] JimBastard_: im just a ruby noob
[16:49] JimBastard_: plus its all sync, so it takes forever to run
[16:49] tjholowaychuk: JimBastard_: oh :) whats the url again? will check it out
[16:49] JimBastard_: needs EM or node
[16:49] tjholowaychuk: haha
[16:49] JimBastard_: http://blog.nodejitsu.com/most-influential-github-users-by-location
[16:49] JimBastard_: i used someone elses code for it, with slight mods
[16:49] JimBastard_: you know how i roll
[16:50] JimBastard_: mikeal and isaacs , i updated article to include oakland
[16:50] JimBastard_: congrats
[16:50] isaacs: aewsome
[16:50] JimBastard_: node users took a bunch of top 5s
[16:50] JimBastard_: tjholowaychuk: http://gist.github.com/608869 and http://github.com/mattb/flotsam/tree/master/github-recruitment/
[16:51] JimBastard_: id be real interested is modding the ruby script to do the changes i mention in the post
[16:51] JimBastard_: my ruby is lacking though
[16:51] JimBastard_: and ive never used EM before
[16:51] trotter has joined the channel
[16:51] JimBastard_: SubStack: wtf man, you changed your LOC
[16:52] creationix: JimBastard: Palo Alto, lol
[16:52] JimBastard_: NOW I HAVE TO RERUN ALASKA
[16:52] JimBastard_: yeah, congrats creationix 
[16:52] JimBastard_: i want to create normaliztions so i can do "california" and it picks up on all those regions
[16:52] JimBastard_: so we can get aggregate ratings
[16:52] slashxr has joined the channel
[16:52] creationix: that would be cool
[16:52] creationix: or bay area
[16:52] JimBastard_: yeah for sure
[16:52] benv has joined the channel
[16:52] JimBastard_: i talk about how to make the changes in the post, maybe ill get around to it sometime
[16:53] javajunky: hey isaacs does npm support private registries atm ?
[16:53] JimBastard_: this tech recruiter i know its all about it, he wants to have a meeting or something, meehhh
[16:53] creationix: if I ever move back to Texarkana (pop 100k) I'll win for sure there ;)
[16:53] isaacs: javajunky: npm config set registry http://somewhere/else/
[16:53] JimBastard_: creationix: winning palo alto is serious business imo
[16:53] JimBastard_: lol
[16:53] javajunky: isaacs: cheers man, you're a star ;)
[16:53] qFox has joined the channel
[16:53] tjholowaychuk: JimBastard_: I question that script when im #2 in canada lol something seems.. off
[16:54] tjholowaychuk: either that or canadians suck
[16:54] slashxr: wheres everyone from?
[16:54] slashxr: im from LA
[16:54] isaacs: slashxr: didn't you know?  i'm from oakland.  http://blog.nodejitsu.com/most-influential-github-users-by-location
[16:54] creationix: JimBastard_: I don't know, seems most business here is closed-source startups
[16:54] JimBastard_: tjholowaychuk: the guy who beat you seems legit? bo
[16:54] isaacs: lolz
[16:54] JimBastard_:  /s/bo/no?
[16:55] tjholowaychuk: JimBastard_: im just suprised there are not people above both of us
[16:55] creationix: JimBastard_: split up Canada by city and tjholowaychuk will win for sure
[16:55] JimBastard_: creationix: yeah
[16:55] JimBastard_: tjholowaychuk: it makes perfect sense to me
[16:55] tjholowaychuk: haha yeah victoria is useless for tech
[16:55] JimBastard_: tjholowaychuk: if you read the post i talk about the flaws
[16:56] jameshome_ has joined the channel
[16:56] JimBastard_: canada is especially affected by the normalization issue
[16:56] JimBastard_: since most people like province 
[16:56] JimBastard_: link*
[16:57] JimBastard_: i would think tjholowaychuk and cloudhead are top 5 in canada as a whole
[16:57] mjsilva has joined the channel
[16:57] javajunky: nahh, they both suck ;)
[16:57] tjholowaychuk: would be interesting to have it even work with commits on other repos like the sencha/learnboost work I do etc
[16:58] creationix: true, connect has a lot of followers
[16:58] JimBastard_: tjholowaychuk: exactly, i talk about the organization problem as well
[16:58] jchris has joined the channel
[16:58] mjr_: Regional github influence is fascinating.
[16:58] JimBastard_: :-)
[16:58] slashxr: cool
[16:58] JimBastard_: i agree!
[16:59] stephenjudkins has joined the channel
[16:59] slashxr: how does that work
[16:59] JimBastard_: it combines my two favorite things: open-source software, and winning!
[16:59] creationix: lol
[16:59] slashxr: i dont understand its purpose
[16:59] JimBastard_: slashxr: to promote nodejitsu
[16:59] JimBastard_: :-D
[16:59] creationix: bloat ego ;)
[16:59] JimBastard_: its a great hiring tool
[16:59] JimBastard_: i know tech recruiters who would kill for this
[16:59] JimBastard_: if it worked a bit better
[17:00] mjr_: The purpose is to be fascinating. 
[17:00] slashxr: so
[17:00] dnolen has joined the channel
[17:00] slashxr: is npm the ruby gems for node?
[17:00] JimBastard_: yep
[17:00] martin-g has joined the channel
[17:01] rauchg_: JimBastard: it indeed has potential 
[17:01] rauchg_: good work
[17:01] JimBastard_: thanks! im surprised you didnt place
[17:01] JimBastard_: i think i saw you in a top 10 list though
[17:01] rauchg_: im in san francisco 
[17:01] benreesman: isaacs in the house?
[17:01] javajunky: JimBastard: is it just based on followers or more clever /
[17:01] rauchg_: too competitive :P
[17:01] JimBastard_: yeah for sure
[17:01] JimBastard_: javajunky: for now its followers / following, which isnt great
[17:01] slashxr: anyone got any live samples of their work done in node.js
[17:02] tjholowaychuk: rauchg_: plus it doesnt link all of your LB work :p
[17:02] tjholowaychuk: w/ socket.io n stuff
[17:02] rauchg_: yep true 
[17:02] JimBastard_: a few people made top 10 just by following lots of people
[17:02] rauchg_: i release most of my code through my company
[17:02] JimBastard_: its not very accurate now
[17:02] benreesman: NPM won't install for me
[17:02] benreesman: http://gist.github.com/610052
[17:02] javajunky: JimBastard:  X/ 0 ? 
[17:02] JimBastard_: javajunky: the code is in the blog post
[17:02] JimBastard_: as well as a write up of the current flaws
[17:02] javajunky: hell I'm changing my location, and un-following, I'd go straight to 2 in England atm ;)
[17:03] javajunky: JimBastard: I did read it when you posted the other day, I'm just *really* lazy ;
[17:03] JimBastard_: me too, its cool
[17:03] benburkert has joined the channel
[17:03] JimBastard_: perhaps its time to do real work today and stop circle jerking
[17:03] JimBastard_: yes, that sounds nice
[17:04] noahcampbell has joined the channel
[17:04] isaacs: benreesman: yep.  http://twitter.com/izs/status/26378278618
[17:04] benreesman: thanks
[17:05] matjas has joined the channel
[17:07] benv has joined the channel
[17:12] charlenopires has joined the channel
[17:17] ooooPsss has joined the channel
[17:18] ollie has joined the channel
[17:19] aheckmann has joined the channel
[17:20] bartt has joined the channel
[17:20] sh1mmer has joined the channel
[17:22] davidsklar has joined the channel
[17:23] fictorial has joined the channel
[17:25] sprout has joined the channel
[17:28] ph^ has joined the channel
[17:29] ph^ has joined the channel
[17:30] ph^ has joined the channel
[17:32] trotter has joined the channel
[17:33] aslakhellesoy has joined the channel
[17:33] aslakhellesoy: Hi folks - anyone knows how far http://github.com/isaacs/npm/issues#issue/266 is from a fix?
[17:33] aslakhellesoy: I can't npm install anything :-/
[17:34] isaacs: aslakhellesoy: yep, looking into it
[17:34] sschuermann has joined the channel
[17:34] aslakhellesoy: hehe, i thought you were. The curse of writing popular software :-)
[17:35] dpritchett has joined the channel
[17:35] technoweenie has joined the channel
[17:35] JimBastard_: hey aslakhellesoy !
[17:35] JimBastard_: << marak
[17:35] JimBastard_: good to see you here!
[17:35] c4milo: aslakhellesoy: you must have it too :) 
[17:35] aslakhellesoy: yeah, i do :-)
[17:36] JimBastard_: aslakhellesoy: random question, can you tell me how to phonetically pronounce your first name?
[17:36] aslakhellesoy: JimBastard_: how is kyuri going?
[17:36] aslakhellesoy: In Norwegian? Not over IRC ;-)
[17:36] JimBastard_: aslakhellesoy: we are focusing on getting nodejitsu publicly launched, indexzero will be getting back on that as soon as we can free him up a bit
[17:37] JimBastard_: prenup has had a bunch of updates too
[17:37] JimBastard_: hopefully will get both stable for use before end of year
[17:37] Mezriss has joined the channel
[17:37] aslakhellesoy: JimBastard_: I met a smart guy the other day who is working on a ruby2js compiler
[17:38] aslakhellesoy: uses rubinus' parser/ast to write out js sources
[17:38] JimBastard_: im just starting to fully understand BDD, im a bit of a BDD and ruby noob, so its been a sharp learning curve
[17:38] tmpvar has joined the channel
[17:38] JimBastard_: that sounds interesting
[17:38] aslakhellesoy: According to himself: "I'm now struggling to write ruby that doesn't translate and execute well"
[17:38] JimBastard_: ahahahahaha
[17:38] rwaldron_ has joined the channel
[17:39] vnguyen has joined the channel
[17:39] rwaldron_ has joined the channel
[17:39] JimBastard_: it would be nice to get some committers to kyuri, i think i saw a gist of yours the other day for gherkin.js
[17:39] rwaldron_ has joined the channel
[17:39] kevwil has joined the channel
[17:39] JimBastard_: if you have any free time (lol) you should try and poke indexzero a bit
[17:40] JimBastard_: i think he has the majority of the work already done, its just a matter of packaging it up 
[17:40] ph^ has joined the channel
[17:41] iganapolsky has joined the channel
[17:42] derferman has joined the channel
[17:42] sschuermann: hello 
[17:44] sschuermann: is the npm registry down? 
[17:44] JimBastard_: aye
[17:44] JimBastard_: he's on it
[17:44] sschuermann: ah ok ;9 
[17:44] sschuermann: thanks 
[17:44] sschuermann: wanted to release 0.0.3 ;) fuzzy rules 
[17:44] sschuermann: maybe a hint that i should work on the defuzzyfication code 
[17:44] ph^ has joined the channel
[17:45] JimBastard_: sschuermann: fuzzy testing?
[17:45] JimBastard_: fuzz testing?
[17:45] sschuermann: no, fuzzy logic 
[17:45] JimBastard_: meh
[17:45] sschuermann: still very basic 
[17:45] ryah: felixge_: yes
[17:45] sschuermann: i want to combine it with a workflowengine, a async one 
[17:45] sschuermann: and use it for gameengines
[17:45] JimBastard_: sschuermann: the only real-world application i know of that uses fuzzy logic is stablization in camcorders
[17:45] JimBastard_: i should research it more
[17:45] ph^ has joined the channel
[17:46] sschuermann: yeah, read the game programming ai book from oreilly 
[17:46] sschuermann: i want fuzzy-statemachines ;) 
[17:46] JimBastard_: im more interested in http://en.wikipedia.org/wiki/Fuzz_testing
[17:47] killfill has joined the channel
[17:47] killfill: hi
[17:47] JimBastard_: yes, its down
[17:47] bradleymeck: yello
[17:47] JimBastard_: sup
[17:47] killfill: anyone can point me to an example how to work with sessions'
[17:47] killfill: ?
[17:47] mbrochh has joined the channel
[17:47] eazyigz has joined the channel
[17:47] mbrochh has joined the channel
[17:47] bradleymeck: which server are you using
[17:47] JimBastard_: killfill: with or without a framework
[17:48] JimBastard_: i use http://github.com/marak/session.js
[17:48] killfill: can it do boths?... (im not sure if i should use express or not for my app)
[17:48] dgathright has joined the channel
[17:48] killfill: i actually never fully understood sesions and cookies.. :P
[17:48] JimBastard_: its pretty easy
[17:48] kriskowal has joined the channel
[17:48] sschuermann: killfill: express == good, better as using just node for webapps 
[17:49] JimBastard_: read that code i sent ya, its pretty simple
[17:49] bruse: the server and the client shares a secret.. as long as the client sends the secret to the server for every request the servers knows the identity of the client
[17:49] bruse: easy as pie
[17:49] killfill: sschuermann: yup, i may use node just for passing json around.. :)
[17:49] JimBastard_: hell yeah
[17:49] JimBastard_: thats a good use
[17:49] sschuermann: killfill: i even use express here too 
[17:49] sschuermann: ;) 
[17:49] prettyrobots has joined the channel
[17:49] killfill: bruse: yup.. exactly what i need.. login validation :)
[17:50] bradleymeck: using node to just pass anything around is generally a good use case for node
[17:50] killfill: JimBastard_: (thanks)
[17:50] JimBastard_: anytime!
[17:50] rwaldron has joined the channel
[17:51] killfill: express has session stuff builded in?
[17:51] felixge_: ryah: sorry, was on the phone
[17:51] felixge_: ryah: do you think the test makes sense, or am I doing it wrong?
[17:52] tjholowaychuk: killfill: yeah, Express is built on connect (basically Rack if you are from ruby-land) which has session support
[17:52] tjholowaychuk: and stores for redis, mongo, etc
[17:53] killfill: hm.. too bad its not 100% documented.. :P  
[17:53] agnat has joined the channel
[17:53] tjholowaychuk: killfill: haha sorry, it is in connect's docs, but yeah I can add some docs to expressjs.com
[17:54] JimBastard_: i heard the guy that wrote express is the second best developer in canada 
[17:54] JimBastard_: id rather use the framework written by the number 1 guy
[17:54] JimBastard_: whats his framework called
[17:55] slashxr has joined the channel
[17:55] JimBastard_: trooolllololo
[17:55] tjholowaychuk: haha :p he seems to only work with EM
[17:55] CIA-77: node: 03Ryan Dahl 07master * r38f8665 10/ (lib/http.js test/simple/test-http-proxy.js): 
[17:55] CIA-77: node: lib/http.js - nested loop stepping on variables of outer loop
[17:55] CIA-77: node: Fixes #318. - http://bit.ly/aqB4Bc
[17:55] tjholowaychuk: outta luck
[17:55] killfill: its cool, becouse im playing with a sencha frontend.. its cool to know connect gets some love from sencha too.. :P
[17:56] c4milo: ahahah 
[17:56] c4milo: I didn't know that !! ? I think I will leave expresso and expressjs 
[17:57] codetonowhere has joined the channel
[17:57] c4milo: :)
[17:57] ryah: felixge_: i just think v8 isn't very good a collecting objectwrap things..
[17:57] c4milo: hehehe
[17:57] ryah: felixge_: but i don't know...
[17:57] felixge_: ryah: well, it seems like it *never* collects them
[17:57] eazyigz: what's the best templating engine for nodejs if I want to display charts/graphs
[17:57] eazyigz: ?
[17:57] felixge_: ryah: at least any memory claimed by buffers isn't being freed right now as far as I can tell
[17:57] felixge_: ryah: and I don't think this has been the case before fast buffers
[17:58] killfill: tjholowaychuk: connect has actually a doc?...  /me cannot find it
[17:58] felixge_: ryah: need to run the test on an old version to be sure so
[17:58] nikolaiwarner has left the channel
[17:58] c4milo: I would say jade but it was wroten by the second best developer in Canada, so we need figure out another one 
[17:58] langworthy has joined the channel
[17:58] tjholowaychuk: killfill: http://senchalabs.github.com/connect/
[17:58] ryah: felixge_: no, memory is almost always freed from buffers
[17:58] csullivan has joined the channel
[17:58] ryah: felixge_: sometimes it lingers
[17:58] killfill: :P
[17:58] paul__ has joined the channel
[17:58] tjholowaychuk: killfill: seems they have not been pushed to gh-pages for a bit
[17:59] tjholowaychuk: seem out of date
[17:59] ryah: felixge_: buffers are used everywhere - we'd be very aware if buffers weren't etting freed
[17:59] stephenjudkins has joined the channel
[17:59] killfill: 
ACTION hopes there were a simple session example around to watch..

[17:59] creationix: tjholowaychuk: "they"?
[18:00] spetrea: if I want to make a beamer presentation with Javascript code how do I convert JS to LaTeX ?
[18:00] tjholowaychuk: killfill: look in the express repo, ./examples/session
[18:00] tjholowaychuk: creationix: haha k, "we"
[18:00] felixge_: ryah: I think there is a problem somewhere. You are running the fast-buffers in production at joyent?
[18:00] creationix: tjholowaychuk: do I need to push it?  I've never done it
[18:01] tjholowaychuk: creationix: cant remember if anything has changed really for them, all i know is the first page there has
[18:01] tjholowaychuk: the big long require for the mem store
[18:01] tjholowaychuk: which is now connect.session.MemoryStore
[18:01] tjholowaychuk: etc
[18:01] creationix: yeah
[18:01] creationix: seems I'm missing dox and npm is still busted
[18:01] tjholowaychuk: shit
[18:02] creationix: and I'm too lazy to install it by hand ;)
[18:02] killfill: tjholowaychuk: thanks!
[18:02] spetrea: ryah: did you ever write a LaTeX beamer presentation ? how did you cope with highlighting js code ?
[18:02] ryah: felixge_: i'm running a lot of tests on fast-buffers
[18:02] herbySk: ryah: the only problems I thought of when saq fast buffers was that one or two fastbuffers holding their data for long time can block the whole holder and if this sums up, significant part of memory may be very inefficiently blocked... but I don;t know if that happens in reality.
[18:02] ryah: felixge_: yeah, there's probably a problem 
[18:02] benv has joined the channel
[18:03] dahankzter has joined the channel
[18:03] ryah: spetrea: off topic
[18:04] ryah: spetrea: (but use the 'listings' package)
[18:04] felixge_: ryah: I'm going to run the test in an old node version now
[18:04] spetrea: ryah: pm ?
[18:04] emmanueloga has joined the channel
[18:05] rcy has joined the channel
[18:06] ryah: spetrea: sure
[18:10] matjas has joined the channel
[18:11] voodootikigod has joined the channel
[18:11] franksvalli has joined the channel
[18:12] aubergine_ has joined the channel
[18:13] c4milo has joined the channel
[18:13] sveisvei has joined the channel
[18:14] hoodoos: guys, I expect a GC problem with nodejs module containg like 10k items in hashes. It begin to throw free() or malloc() errors. Currently I'm trying to implement a test case module to repeat this issue, is there any way to call GC forcibly?
[18:14] hoodoos: *expirience :)
[18:15] hoodoos: *module is written in javascript
[18:15] hoodoos: *node version is 0.2.0
[18:15] wink_: hoodoos: i've used far larger arrays without issue
[18:16] bradleymeck: a: which module, b: what kind of hash
[18:16] hoodoos: let me post a code for it
[18:16] ryah: hoodoos: do "node --expose_gc" and call "gc()" in your code
[18:16] hoodoos: ryah, thanks!
[18:16] galaxywatcher has joined the channel
[18:17] hoodoos: http://pastebin.com/g4rVNL0n
[18:17] hoodoos: here's code
[18:18] eazyigz has joined the channel
[18:18] ryah: looks like you're using libxmljs
[18:18] wink_: hoodoos: what are the actual free/malloc errors?  double frees?
[18:18] ryah: so it's not pure js
[18:18] wink_: because libxmljs is almost definitely at fault
[18:18] hoodoos: year, but it works find with init
[18:19] hoodoos: but, later errors appear
[18:19] ryan[WIN]: json 4 lyfe
[18:19] c4milo has joined the channel
[18:19] hoodoos: wink_, yeah, i know it's somewhat buggy if not used as SAX parser..
[18:19] aubergine has joined the channel
[18:20] wink_: hoodoos: just for a point of reference, i've filled an array with 350mb worth of mongo documents (~43000) with no issue
[18:21] hoodoos: and if you'll make it hash? I mean object with 43000 fields? It's literaly the same in JS i guess
[18:21] bradleymeck: hoodoos it seems libxmljs isnt using makeweak which leads me to believe there could be an issue
[18:21] wink_: yeah, each array element was another set of ~20 fields some of which were arrays of arrays
[18:22] franksvalli has joined the channel
[18:22] bradleymeck: as a side note the v8 identity hash has around 1/50,000 colision rate
[18:22] felixge_: ryah: ok, seems like this is not a new problem. Pre-fastbuffer node behaves the same way
[18:23] hoodoos: wink_, bradleymeck, http://groups.google.com/group/nodejs/browse_thread/thread/742b73134eca9145 here's traces of some errors
[18:23] ysynopsis1 has left the channel
[18:23] hoodoos: wink_, it actually lists all c modules I use in it :)
[18:24] [[zz]] has joined the channel
[18:24] hoodoos: node_compress is not an option because it works perfectly if I don't require this particular module
[18:25] wink_: that may or may not be true though, it could easily stomp on something that isnt noticed until you start sucking up address space
[18:25] hoodoos: wink_, yeah, I see when it actually comes to free addresses?
[18:25] dilvie has joined the channel
[18:26] wink_: hoodoos: are the free/malloc errors random?  i see your valgrind trace is reporting a malloc failure not a free
[18:26] wink_: is that from execution of the same code?
[18:26] dilvie1 has joined the channel
[18:27] hoodoos: wink_, yeah, they're random, sometimes it's corruption, sometimes malloc/free or sometimes mix one after another
[18:27] csullivan has joined the channel
[18:27] wink_: and valgrind isnt reporting out of bounds writes or anything?
[18:27] hoodoos: wink_, it reports that everything is fine, I can try to run it again, i'm not very familiar with it
[18:27] CrabDude has joined the channel
[18:28] isaacs has joined the channel
[18:28] shockie has joined the channel
[18:28] tilgovi has joined the channel
[18:29] wink_: run it with valgrind --leak-check=full --error-limit=no
[18:29] wink_: that sort of corruption should get picked up by valgrind pretty easily
[18:30] wink_: also, http://github.com/polotek/libxmljs is that the lib you're using?
[18:31] joshteam has joined the channel
[18:31] hoodoos: wink_, yes
[18:31] hoodoos: wink_, among others
[18:31] hoodoos: *with
[18:31] aubergine_ has joined the channel
[18:31] joshteam: any good resources on setting node.js up on deb (lenny) ?
[18:32] hoodoos: joshteam, development or production?
[18:32] joshteam: hoodoos: dev for now.
[18:32] hoodoos: joshteam, just clone and install it :)
[18:32] Druid_ has joined the channel
[18:32] JimBastard_: joshteam: it should be easy? just clone and install? i think i saw a .deb package on the mailing list
[18:33] hoodoos: wink_, damn, it works fine by now..
[18:33] wink_: hoodoos: lets see what we can do about valgrind, i think thats probably the easiest way to go
[18:34] wink_: run it outside of valgrind, is it still crashing there?
[18:34] hoodoos: let me try
[18:34] hoodoos: the most scarry i can't remember what i was upgrading last few days..
[18:34] V1: OMG I GOT A STABLE SOCKET.IO@@@ O_O
[18:34] V1: Yes that message needed capslocks
[18:35] hoodoos: wink_, valgrind show no frees, maybe that's why it won't crash
[18:35] hoodoos: wink_, let me try to force call GC
[18:35] wink_: try to get your script to exit ?
[18:35] wink_: it should at least cleanup then
[18:35] shockie has joined the channel
[18:35] c4milo has joined the channel
[18:36] hoodoos: wink_, no errors on exit
[18:36] LFabien has joined the channel
[18:36] wink_: sweet you fixed it ;D
[18:36] stephenjudkins has joined the channel
[18:36] hoodoos: wink_, no i'm not satisfied yet, it won't me live I know..
[18:36] paul__ has joined the channel
[18:37] wink_: haha
[18:37] wink_: well get it broken again and we'll try again later
[18:37] eazyigz: how do I return an html page to the client?
[18:37] JimBastard_: eazyigz: lol
[18:37] JimBastard_: response.write('fuuuuuckkk')
[18:37] eazyigz: I know res.writeHead(200, ....)
[18:38] JimBastard_: text/html is the content type you want
[18:38] isaacs: eazyigz: response.write("<html>hi there")
[18:38] JimBastard_: you gotta set content type
[18:38] shockie has joined the channel
[18:38] JimBastard_: well, you should
[18:38] eazyigz: but I want to return an already created html page
[18:38] JimBastard_: eazyigz: reverse proxy?
[18:38] JimBastard_: file system ?
[18:38] eazyigz: isaacs: I mean return an html document, not html code
[18:39] JimBastard_: o.O
[18:39] JimBastard_: you mean read a .html from the file system?
[18:39] eazyigz: I don't want to create html tags in my nodejs code
[18:39] eazyigz: I already have the html page, I just want to display it
[18:39] hoodoos: ryah, is it right way to start node? node --expose_gc ../../src/mcore/core.js config.json. it won't allow me to call gc() says she doesn't know one
[18:39] JimBastard_: eazyigz: use node-static
[18:39] JimBastard_: eazyigz: http://github.com/cloudhead/node-static
[18:39] hoodoos: ryah, btw, i do it from child process, maybe that's why?
[18:40] konobi: hoodoos: you can't do GC calls from node
[18:40] admc has joined the channel
[18:40] paul__ has joined the channel
[18:40] isaacs: eazyigz: http = require("http") ; fs = require("fs") ; http.createServer(function (req, res) { fs.readFile("index.html", function (er, s) { if (er) { res.writeHead(404, {}) ; res.end("not found") ; return } res.writeHead(200, {"content-type" : "text/html", "content-length" : s.length}) ; res.end(s) }})
[18:40] dpritchett: i use the static component of Connect:               Connect.staticProvider(__dirname + '/public'),
[18:40] hoodoos: konobi, well, maybe i got it wrong :)
[18:40] JimBastard_: isaacs: thats not really a great idea, no caching
[18:40] JimBastard_: FD limit + long response times
[18:40] konobi: infact calling GC from C++/v8 space is pretty damn tricky
[18:41] hoodoos: konobi,  hoodoos: do "node --expose_gc" and call "gc()" in your code. he meant c not js?
[18:41] isaacs: JimBastard: he asked how to return an html doc
[18:41] isaacs: JimBastard: not how to return an html doc WELL
[18:41] JimBastard_: true life
[18:41] hoodoos: konobi, it what's ryah said
[18:41] JimBastard_: of course
[18:41] isaacs: eazyigz: for better results, cache the data for some amount of time, etc.
[18:41] konobi: hhhmmm... must be new
[18:41] creationix: sendfile!
[18:41] isaacs: sure, that too
[18:41] JimBastard_: anytime i write a line of node.js code these days, the first thing that comes to my mind is, how well will this work on the front page of HN
[18:42] saikat has joined the channel
[18:42] isaacs: JimBastard: that site totally trolls you
[18:42] wink_: lol
[18:42] JimBastard_: troolololo
[18:42] JimBastard_: PG can suck it
[18:42] isaacs: hah
[18:42] isaacs: it's not PG that i mind.  it's everyone else on HN
[18:42] JimBastard_: as much as i like my investors to believe there is an 80% chance ill fail
[18:43] eazyigz: isaacs: to cache results do I need to use a buffer or something?
[18:43] JimBastard_: you could just cache in memory with an object
[18:43] hoodoos: konobi, it doesn't work anyways :) atleast in child process, let me test it in parent..
[18:43] isaacs: eazyigz: wel, what you should do is: stat the file, and keep the stat object (or the error) in memory. then, when you get the request, map it to one of the cached stat files, open it, and sendfile the fd to the res.connection.fs
[18:43] isaacs: *fd
[18:43] JimBastard_: basic http cache, http://github.com/nodejitsu/node-http-cache/blob/master/lib/http-cache.js
[18:44] isaacs: eazyigz: or just use paperboy or something
[18:44] isaacs: eazyigz: that's probably WAY easier and betterer
[18:44] JimBastard_: *cough* node-static
[18:44] isaacs: sure, or that
[18:44] isaacs: node-static
[18:44] JimBastard_: :p
[18:44] isaacs: 
ACTION hands JimBastard a lozenge

[18:44] JimBastard_: i gotta respect the top developer in oakland
[18:45] hoodoos: wink_, got it! =)
[18:45] hoodoos: wink_, but with no valgrind (
[18:45] wink_: to break?
[18:45] hoodoos: wink_, yea
[18:45] dpritchett: ##javascript hurts my feelings more than jslint
[18:45] wink_: ah does it not break with valgrind?
[18:46] shockie has joined the channel
[18:46] hoodoos: wink_, can't catch it yet
[18:46] stephenjudkins has joined the channel
[18:46] alcuadrado has joined the channel
[18:46] hoodoos: wink_, error was as usual: *** glibc detected *** /usr/local/bin/node: malloc(): memory corruption: 0x0000000002886460 ***
[18:47] hoodoos: wink_, does it mean anything if it won't fire with valgrind?
[18:47] ngw has joined the channel
[18:47] Vekz has joined the channel
[18:48] MikhX has joined the channel
[18:48] nefD: whoa.. chrome (dev channel) has a 'web socket' tab in the developer tools now
[18:48] dpritchett: neat
[18:49] hoodoos: wink_, let me leave it for some time..
[18:49] alcuadrado: has anyone tested the performance implication to use in-line style versus classes (lots of classes, for lots or elements)
[18:49] alcuadrado: ?
[18:50] rnewson has joined the channel
[18:50] wink_: hoodoos: yeah it probably does, unfortunately no tellign what ;>
[18:50] wink_: see if you can increase your dataset size
[18:50] bradleymeck: alcuadrado jsperf.com
[18:50] wink_: even artificially
[18:51] alcuadrado: bradleymeck, thanks, I'll take a look at it
[18:52] mischievious has joined the channel
[18:56] nichdiekuh has joined the channel
[18:56] mischief__ has joined the channel
[18:59] Nohryb has joined the channel
[18:59] kevwil has joined the channel
[19:00] mischief has joined the channel
[19:03] pkrumins: cake or death?
[19:03] JimBastard_: CAKE
[19:03] JimBastard_: plz
[19:04] pkrumins: well, we're out of cake
[19:04] pkrumins: but we have chicken
[19:04] kgf: I'll have the chicken then
[19:04] JimBastard_: i want to live!
[19:04] JimBastard_: sorry, we are fresh out of that
[19:04] pkrumins: alright, there you go
[19:04] kgf: tastes of human, sir.
[19:04] JimBastard_: sorry, we are fresh out of that
[19:05] pkrumins: in any case thanks for flying the church of england.
[19:05] csullivan_ has joined the channel
[19:07] spetrea: I want to make a node.js presentation
[19:07] ginader has joined the channel
[19:07] creationix: spetrea: go for it
[19:08] spetrea: so far I've settled on my latex environment because ryah was kind enough to give me his .tex
[19:08] spetrea: now I have the following
[19:08] spetrea: I will present   node.js an echo server, npm, Faye
[19:08] spetrea: I don't know what else to include
[19:08] spetrea: the presentation is supposed to last maybe at most 1h
[19:09] peritus- has joined the channel
[19:09] codetonowhere has joined the channel
[19:10] JimBastard_: spetrea: you should mention connect i think
[19:10] JimBastard_: for middleware
[19:10] dpritchett: socket.io is fun
[19:10] peritus-: wink_: can you look at http://dl.dropbox.com/u/8735936/valgrind.log ? that's my malloc fault from a few hours ago.
[19:10] dpritchett: i am happy with my little chatroom that receives json from clients and then broadcasts it back to all of them simultaneously
[19:10] JimBastard_: yeah both of those
[19:10] spetrea: JimBastard_: link pls ?
[19:10] JimBastard_: connect and socket.io
[19:11] spetrea: dpritchett: I tried socket.io but Faye worked better for me so I would like to present Faye instead
[19:11] JimBastard_: github.com/senchalabs/connect
[19:11] spetrea: dpritchett: I just couldn't get socket.io to work :(
[19:11] wink_: peritus-: sure one sec
[19:11] spetrea: Faye and Socket.IO are doing the same thing
[19:11] dpritchett: sure spetrea i understand completely
[19:11] dpritchett: it took me a few hours to figure out what i was doing withs socket.io but i'm pretty happy with it now
[19:12] peritus-: wink: that's with node 0.2.3 and mac os 10.6.4
[19:12] ooooPsss: does Pusherapp uses socket.io ?
[19:12] wink_: peritus-: what was your crash again?
[19:13] cloudhead has joined the channel
[19:13] peritus-: node_g(49636) malloc: *** error for object 0x100c25608: incorrect checksum for freed object - object was probably modified after being freed.
[19:13] peritus-: *** set a breakpoint in malloc_error_break to debug
[19:13] peritus-: Abort trap
[19:13] iganapolsky has joined the channel
[19:14] wink_: peritus-: ok one sec
[19:14] itissid: isaacs: Hey npm is still down
[19:14] isaacs: itissid: i'm aware
[19:14] itissid: Someone suggested mirroring
[19:14] itissid: the couch server.. Its on EC2...
[19:14] itissid: Is that doable?
[19:15] wink_: peritus-: odd, i dont see the crash on ubuntu, one sec
[19:15] isaacs: itissid: sure.  go nuts.  just replicate the couchdb.  but that won't work atm, because it's donw.
[19:15] isaacs: *down
[19:15] isaacs: so, replicating out the attachments will fail
[19:15] voodootikigod: mikeal: 
[19:15] itissid: :(
[19:15] isaacs: since that also hits the view server.  mikeal has looked into it somewhat, but we're waiting for hte dude who can log into that machine
[19:15] isaacs: itissid: good news: the data is fine.
[19:15] isaacs: itissid: afaict
[19:16] JimBastard_: isaacs: im thinking you should code in a seconary end point for npm
[19:16] JimBastard_: so its assumed the primary will go down, like how dns works for most hosting providers 
[19:16] isaacs: JimBastard: sure.
[19:16] JimBastard_: :-)
[19:17] wink_: peritus-: it does seem to leak like a sieve though
[19:17] JimBastard_: you know your open-source project is doing well when it goes down for a little bit and people start freaking out :-D
[19:17] nroot7 has joined the channel
[19:18] JimBastard_: isaacs: should i create issue?
[19:18] isaacs: yeah, seriously! :D
[19:18] JimBastard_: re: secondary provider 
[19:18] isaacs: i think there's an issue already
[19:18] zmbmartin has joined the channel
[19:18] isaacs: to have npm support multiple registries in a list or somethign
[19:18] zmbmartin has left the channel
[19:18] isaacs: just need to do some javascript.  maybe i'll put that in the next version
[19:19] pashields has joined the channel
[19:19] konobi: flatfiles++
[19:19] v8bot:  konobi has given a beer to flatfiles. flatfiles now has 1 beers.
[19:19] itissid: isaacs: So I assume u know who is manageing the server and I assume you fired a mail to him ? :)
[19:19] konobi: v8bot: die!!!
[19:19] v8bot: konobi: Use v8: <code> to evaluate code or "`v commands" for a list of v8bot commands.
[19:20] JimBastard_: http://github.com/isaacs/npm/issues/#issue/100
[19:20] JimBastard_: commented
[19:20] itissid: JimBastard_: Cool..
[19:23] delapouite has joined the channel
[19:24] isaacs: yep
[19:25] isaacs: JimBastard: my plan with that was always to have some system where any registry GET will go down the list in the event of failure.
[19:25] matthewford has joined the channel
[19:25] isaacs: JimBastard: so, it'll GET from the first, thne the second, etc.
[19:25] isaacs: JimBastard: or maybe some kind of way to specify multiple registry.get urls and multiple registry.put urls, i'm not sure
[19:25] wink_: peritus-: can you run it in gdb and get a stack trace?
[19:26] spetrea: npm ! Error: ucs {bad_utf8_character_code}: connect
[19:26] spetrea: wha's going on ?
[19:26] spetrea: why do I get this ?
[19:26] spetrea: I just hit    npm install connect
[19:26] dpritchett: because it's down
[19:26] isaacs: spetrea: known.  looking into it.  couchdb error.
[19:26] spetrea: as root
[19:26] spetrea: isaacs: ok 
[19:26] isaacs: spetrea: not your fault
[19:27] peritus-: wink_: i can try to do that
[19:27] peritus-: wink_: give me 10 min
[19:27] spetrea: I know CPAN has many servers, doesn't npm have mirrors ?
[19:27] wink_: kk
[19:27] itissid: spetrea: :( no
[19:27] isaacs: spetrea: cpan is also 14 years older than npm, and has more than one maintainer.  so, yeah, that's totally planned, but not there yet.
[19:27] wink_: peritus-: just run gdb node_g then run script.js, once it crashes pastebin the output of bt
[19:27] hoodoos: wink_, i got it with valgrind, but it reports no errors
[19:28] hoodoos: wink_, and yes, maybe it's really related with some c libs i use.. i guess i need to spend some time with dbg..
[19:28] wink_: hoodoos: ouch...thats no good
[19:28] aheckmann has joined the channel
[19:28] wink_: hoodoos: yeah the first step needs to be to isolate it to one of the libs
[19:28] hoodoos: wink_, it's hard, it won't repeat all the time at my will :(
[19:29] hoodoos: wink_, btw, ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 4 from 4), it suppressed something
[19:29] wink_: yeah memory bugs are like that :>, there is a flag to unsupress those
[19:29] wink_: see what those look like
[19:29] wink_: also try increasing your dataset size some more
[19:30] hoodoos: i'll load it twice or thrice..
[19:31] kschzt has joined the channel
[19:31] olegp has joined the channel
[19:31] programble has joined the channel
[19:33] hoodoos: wink_, btw, what compress lib for gzip u use, if you use one? maybe I'll try to replace it with other one it will magicaly get fixed
[19:33] rex_fernando has joined the channel
[19:33] wink_: hoodoos: i've not needed one yet so im not even sure what's out there unfortunately
[19:34] hoodoos: thanks for all your help guys :)
[19:34] deepthawtz has joined the channel
[19:36] wink_: hoodoos: one other thing to maybe try is run node with --trace_gc, if it might be the gc eating you, you may be able to spot it when the gc executes
[19:37] hoodoos: okay, i'll give it a shot
[19:38] peritus-: wink_: http://pastebin.com/a01RgFaQ
[19:38] hoodoos: wink_, don't i need some debug build of node?
[19:38] wink_: hoodoos: i dont think you do for --trace_gc
[19:39] wink_: peritus-: execute 'bt' after the crash
[19:39] hoodoos: wink_, yeah, it shows some scavanges :)
[19:39] wink_: hoodoos: yeah i'd expect that, it'd be tough to actually correlate it though :>
[19:39] JimBastard_: AHAHA wow
[19:39] JimBastard_: check this out
[19:39] JimBastard_: in my refferal log
[19:39] JimBastard_: http://habrahabr.ru/blogs/development/105499/
[19:39] JimBastard_: when translated
[19:40] JimBastard_: The current code is written in Ruby and I think the next version will use the event-driven car to speed synchronous requests. When do I get more accurate results - I will publish the entire list for each region. If you want to try the script yourself, you can try it here or here
[19:40] JimBastard_: then the guy puts an update saying
[19:40] JimBastard_: UPD 01:  IamFM and kmike prompted the inaccuracy of the translation of the event-driven car. " Possible (or likely), the author meant the library: statemachine or EventMachine.  In the same sentence mention about node.js I trimmed. 
[19:40] wink_: hoodoos: you may want to try running inside gdb and dumping a back trace as well
[19:40] peritus-: wink_: http://pastebin.com/gKe9RLun
[19:40] JimBastard_: thats really funny
[19:40] JimBastard_: i should have included link
[19:40] wink_: peritus-: k one sec
[19:41] hoodoos: wink_, yeah, ry suggested that, and he also thinks that it's node-compress :)
[19:41] wink_: peritus-: comment out the digest, does it still crash?
[19:41] aheckmann has joined the channel
[19:41] peritus-: wink_: i think that triggers it, but i'll double-check
[19:42] wink_: it sure looks like it does ;>, not that thats super helpful for you
[19:42] wink_: hoodoos: yeah i saw that in the google thread, if you can get a back trace out of gdb though that can go a long way to finding it
[19:44] pedrobelo has joined the channel
[19:44] peritus-: wink_: it's the digest: http://pastebin.com/bG72f5iV
[19:44] hoodoos: wink_, never used dbg, would be nice to have that expirience, i guess when u write with node those tools got to be handy
[19:44] peritus-: wink_: s/digest/.digest()/g
[19:45] wink_: hoodoos: run gdb node_g , you'll get a prompt type 'run script.js' , once it crashes type 'bt' and pastebin that
[19:45] ben_alman has joined the channel
[19:45] hoodoos: wink_, btw, process is not dying after that error, it works fine :))
[19:45] bruse: does node segfault?
[19:46] bruse: oh
[19:46] TomsB has joined the channel
[19:46] hoodoos: bruse, u can make it segfault in several ways :))
[19:46] wink_: never!
[19:46] hoodoos: oops.. yeah, never!
[19:46] hoodoos: :)
[19:46] ajsie has joined the channel
[19:46] derferman has joined the channel
[19:47] peritus-: for large amounts of "never" ;)
[19:47] wink_: peritus-: for some reason it looks like ssl isnt init'ing the ctx datastructure properly
[19:47] bruse: anyway, if you're hunting down a crashing bug, you dont have to have gdb attached
[19:47] bruse: just make it core dump, then load the core dump into gdb
[19:47] wink_: you definitely dont have to
[19:48] hoodoos: bruse, it doesn't crash at all
[19:48] hoodoos: wink_, look at this one http://github.com/egorich239/node-compress/commit/b157081ac257d82fb2d7ef98a5dd378404b26670
[19:48] hoodoos: wink_, looks like my patient :))
[19:48] hoodoos: wink_, *maybe*
[19:48] wink_: are you running the same rev?
[19:48] wink_: or is that an update for you
[19:48] hoodoos: now yes, i just upgraded..
[19:48] wink_: and no more crash?
[19:48] hoodoos: can't tell for sure but yet it didn't crash
[19:48] peritus-: wink_: i think i'm on openssl 1.0.0a .. should I upgrade ?
[19:49] wink_: peritus-: to be honest im not real sure...i cant get it to crash here.  i know ssl support has been a little shaky in the past im not sure thats still the case though
[19:49] nrub has joined the channel
[19:50] wink_: i dont have access to osx to hunt it any further, if you want some line numbers to set break points i can probably help you look a little deeper
[19:50] seanouimet has joined the channel
[19:50] hoodoos: what case u need to make it crash? i consume a lot of services over ssl with http.client, didn't have any problems since 0.1.9..
[19:50] hoodoos: wink_, or is it some special case? :)
[19:50] peritus-: wink_: hm, macports doesn't have an openssl > 1.0.0a
[19:50] micheil: isaacs: I take it npm's not back up yet?
[19:50] wink_: hoodoos: it shouldnt be, hes creating an hmac and calling digest
[19:51] wink_: getting a crash way down in the guts
[19:51] wink_: peritus-: what version of node? 0.2.0?
[19:51] wink_: or newer?
[19:51] peritus-: wink_: 0.2.3
[19:52] csullivan has joined the channel
[19:52] hoodoos: wink_, well, i'll make a fake data provider for my service and leave it for night that will tell me for sure..
[19:52] peritus-: wink_: but IIRC the problem existed on my machine for quite some versions..
[19:52] peritus-: wink_: i can check if you like
[19:53] wink_: peritus-: well the only reason i ask is the version on my disk may not line up line number wise, but if you get a bp set in node_crypto.cc line 1646 (0.2.0), thats where it starts calling into ssl
[19:54] rauchg__ has joined the channel
[19:54] peritus-: wink_: node --version prints "v0.2.3" ..
[19:55] hoodoos: ls
[19:55] hoodoos: damn, i need to go to sleep. thanks again))
[19:55] peritus-: wink_: i'm building node in a subfolder of my project - but it could link to some old lingering global installed librar
[19:55] peritus-: *y
[19:56] peritus-: wink_: what's the equivalent of 'ldd' on darwin ?
[19:56] wink_: sure....if i had a mac i could probably help more , but without access to one im kinda blind :p
[19:56] richcollins has joined the channel
[19:56] wink_: i would've expected it to be ldd
[19:56] wink_: :P
[19:56] peritus-: heh
[19:56] wink_: although doesnt osx use .dynlib or some nonsense
[19:57] killfill: im with express, and have set req.session.logged=true. Next i wish to redirect to some pages in /public. How do i 'protect' the public pages validating that req.session.logged is true, and if not, redirecting to the login page?
[19:59] killfill: tjholowaychuk: how would one do that? :)
[19:59] CIA-77: node: 03isaacs 07master * r8ff7954 10/ (lib/fs.js test/simple/test-fs-readfile-empty.js): Fix #325. Add test and check for zero-length file contents in fs.readFileSync - http://bit.ly/98vDIJ
[19:59] CIA-77: node: 03Ryan Dahl 07master * rdbf0d6b 10/ src/node_object_wrap.h : ClearWeak on ObjectWraps. I /think/ this is the correct semantics - http://bit.ly/bpYUVX
[20:01] javajunky has joined the channel
[20:01] tjholowaychuk: killfill: those files are not so public then are they :p
[20:01] TomsB has joined the channel
[20:02] tjholowaychuk: killfill: for private files id suggest either 1) mount a staticProvider via app.use('/private', express.staticProvider())
[20:02] killfill: thats actually right.. :P.. but i guess it a bad idea to fsRead them one by one, and send them throught the server?
[20:02] tjholowaychuk: and also have a middleware above that for authorization or whatever you need
[20:02] tjholowaychuk: or use res.sendfile()
[20:02] javajunky: like connect-auth ;) 
[20:02] killfill: .sendfile would be innefficient right?..
[20:03] Me1000 has joined the channel
[20:03] tjholowaychuk: killfill: not really no, if needed you can apply caching
[20:04] tapwater has joined the channel
[20:04] peritus-: wink_: checked with dtrace, doesn't open any weird libraries.. pastebin follows
[20:04] zemanel has joined the channel
[20:04] killfill: hmm
[20:04] aliem has joined the channel
[20:05] tjholowaychuk: killfill: res.sendfile() is more suited for things that may be large, supports Range etc, staticProvider is mainly aimed at site assets like js / css etc
[20:05] peritus-: wink_: http://pastebin.com/8CUDdrdp
[20:05] wink_: yeah looks ok
[20:06] wink_: i'd go ahead and set some bps in the HmacInit function and make sure everything looks sane
[20:06] wink_: unless you wanna give me a shell
[20:07] killfill: tjholowaychuk: whats the idea behind the first you told?.. can i put somehow an if in /private previous the staticProvider serves the files?
[20:07] peritus-: wink_: i'll look into setting the bps tomorrow - thanks for the help, i'll keep you posted.
[20:07] wink_: kk
[20:08] wink_: HmacInit in node_crypto is where you'll want to start
[20:08] wink_: start in the static v8 version and move deeper from there
[20:08] tjholowaychuk: killfill: you can "mount" middleware to a path segment. kinda confusing but
[20:09] wink_: peritus-: i dont see how it could happen but maybe something simple like the len is getting clobbered or something
[20:09] tjholowaychuk: killfill: for example if you have staticProvider with a root path to ./public/ and "mounted" via app.use('/public', ..) then /public/style.css will map to ./public/style.css
[20:09] tjholowaychuk: otherwise the default mount to "/" will try /style.css as ./public/style.css
[20:10] killfill: hmm
[20:10] felixge_: ryah: should that last commit effect the gc?
[20:10] wink_: peritus-: there is a return value in the instance HmacInit not being checked, there may be something there as well (HMAC_Init returns an int, it make be returning something useful)
[20:10] charlenopires has joined the channel
[20:12] peritus-: wink_: i always wanted to get into serious c++ debugging ;)
[20:12] wink_: peritus-: one more thing, i just read this in the man pages: http://www.openssl.org/docs/crypto/hmac.html , says HMAC_Init is deprecated after 0.9.6, 1.0.0 could've jacked it up
[20:13] killfill: hm.. maybe its simplier if one could put a hook in the beginning of staticProvider
[20:13] wink_: peritus-: a downgrade may fix it for you
[20:14] charlenopires_ has joined the channel
[20:14] wink_:        HMAC_Init_ex() initializes or reuses a HMAC_CTX structure to use the
[20:14] wink_:        function evp_md and key key. Either can be NULL, in which case the
[20:14] wink_:        existing one will be reused. HMAC_CTX_init() must have been called
[20:14] wink_:        before the first use of an HMAC_CTX in this function. N.B. HHMMAACC_IInniitt(())
[20:14] wink_:        had this undocumented behaviour in previous versions of OpenSSL -
[20:14] wink_:        failure to switch to HHMMAACC_IInniitt_eexx(()) in programs that expect it will
[20:14] wink_:        cause them to stop working.
[20:14] tjholowaychuk: killfill: not really, this applies to all middleware, not just staticProvider
[20:14] wink_: ah crap, sorry guys
[20:14] aconbere has joined the channel
[20:14] tjholowaychuk: killfill: just think of it like legos :)
[20:15] killfill: 
ACTION trying.. :P

[20:15] wink_: peritus-: it looks like they expect it to fail, in fact all the others use the proper _ex version...HMAC probably just needs touching up...
[20:15] wink_: verify by downgrading your ssl lib though
[20:16] charlenopires has joined the channel
[20:16] micheil: is there any way to find out how many sockets a node process is currently holding open?
[20:17] wink_: micheil: lsof maybe?
[20:17] micheil: lsof?
[20:17] wink_: ^ :>
[20:17] tjholowaychuk: micheil: list of open files
[20:17] tjholowaychuk: or soemthing
[20:17] wink_: yeah i think thats the acronym, it'll tell you open fds by process
[20:18] micheil: now to work out how to use it.
[20:18] wink_: and in the case of sockets, it'll give you a little more info
[20:18] pandark_ has joined the channel
[20:18] wink_: micheil: try lsof -p <pid>
[20:18] wink_: i dont remember how to tell it sockets only, but you could always grep those away
[20:20] killfill: tjholowaychuk: thats something of connect right?.. i should search its docs then
[20:20] hoodoos: micheil, hey there, i couldn't yet beat problem with browsers keep-alive against my will :))
[20:20] charlenopires_ has joined the channel
[20:21] tilgovi has joined the channel
[20:21] tjholowaychuk: killfill: yup, it is just not used overly often so it doesnt get tons of attention
[20:21] tjholowaychuk: but potentially pretty useful
[20:22] tjholowaychuk: fab handles it in a cooler way, but even more confusing and somewhat overkill lol
[20:22] killfill: yah.. i kind of not find much docs..
[20:22] killfill: guess its like "pipe & messages" pattern.. :P
[20:25] derferman has joined the channel
[20:32] jchris has joined the channel
[20:32] abiraja has joined the channel
[20:33] itissid: isaacs:  here is a thought... Why cant you host the registry on Joyent or chunkhost... They have free start up and 20GB disk with 30GB or so B/W 512MB Ram and a debian machine... I mean i dont know the registry size or the traffic is on the couch server.. But if you think its within reasonable limits you could host no..?
[20:33] isaacs: itissid: also, i work at joyent.
[20:33] isaacs: :)
[20:33] itissid: :)
[20:34] isaacs: so, yeah, absolutely
[20:34] admc has joined the channel
[20:34] mjsilva has left the channel
[20:34] creationix_ has joined the channel
[20:36] digitalspaghetti: so isaacs these no.de instances, any idea of what the cost is going to be?  reason is i'm just running a small site off it, so hoping there will be a small package :D
[20:37] JimBastard_: digitalspaghetti: don't worry, whatever they charge we will go cheaper :-D
[20:37] dshaw has joined the channel
[20:38] killfill: oh i think i got it.
[20:38] killfill: tjholowaychuk: thanks!
[20:38] tjholowaychuk: np
[20:42] WALoeIII has joined the channel
[20:43] aheckmann has joined the channel
[20:46] derferman has joined the channel
[20:47] killfill: amazing :)
[20:47] mikeal1 has joined the channel
[20:49] daleharvey has joined the channel
[20:53] itissid: Hey so if NPM is down how do i roll out wheat on my server with relatively less pain..?
[20:54] isaacs: itissid: download the git repo, cd into it, and do "npm install"
[20:54] isaacs: itissid: install its dependencies first
[20:54] isaacs: (the same way)
[20:54] creationix_: itissid: what he said
[20:54] creationix_: and wheat has a LOT of deps
[20:54] creationix_: but they're all listed in the package.json file
[20:54] creationix_: download + npm install in the folder
[20:55] itissid: Hmm thanks
[20:56] tilgovi has joined the channel
[20:57] isaacs: ah!
[20:57] isaacs: it appears to be up again!
[20:57] itissid: !
[20:57] charlenopires has joined the channel
[20:57] itissid: We have lift off!
[20:57] daleharvey: yup is working for me
[20:59] jchris has joined the channel
[20:59] MikhX has joined the channel
[20:59] creationix_: anybody have a node binary for v0.2.3 for snow leopard
[20:59] halfhalo: binary? psh
[21:00] creationix_: I'm on a different laptop today and don't really have time to download 3gb of xcode just to compile node
[21:00] creationix_: slow network connection
[21:00] creationix_: 
ACTION wishes ivy-bin was up to date

[21:00] benreesman: sorry if this question has been answered already, but is npm back up?
[21:00] tjholowaychuk: creationix_: i have a 0.2.2 but thats it
[21:01] creationix_: tjholowaychuk: good enough
[21:01] creationix_: I just need the node binary, nothing else
[21:01] creationix_: it's nice and portable now ;)
[21:01] tjholowaychuk: creationix_: should be in http://github.com/visionmedia/nodes
[21:01] isaacs: benreesman: yeah, back up
[21:01] benreesman: as usual you rule
[21:02] creationix_: tjholowaychuk: says i686, is that for linux or osx?
[21:02] isaacs: nope, all glory to the couchdb folks and davisp
[21:02] creationix_: default on osx is 64 bit
[21:02] benreesman: btw do you hang with the basho guys at all?
[21:02] tjholowaychuk: creationix_: nvm I only have 0.2.1
[21:02] benreesman: they're involved with joyent too i think arent they?
[21:03] ph^ has joined the channel
[21:04] creationix_: meh, is there something smaller than xcode that can compile node on osx?
[21:06] blowery_zzz has joined the channel
[21:06] bradleymeck: gcc has worked fine to my knowledge
[21:08] isaacs: creationix: xcode is the fucking bane of os x
[21:08] isaacs: creationix: "Oh, you want a C compiler?  DOWNLOAD A BILLION GIGS OF IPAD CRAP SDK BULLSHIT!!! MWAHAHAHAHA!!!!"
[21:09] tjholowaychuk: isaacs: and when you try to update it, they require snow leopard haha
[21:09] tjholowaychuk: assholessss
[21:10] bradleymeck: xcode is nice as long as you dont want to change anything about the default settings, then it chains on forever
[21:10] sschuermann: npm works again ;) 
[21:10] sschuermann: thanks
[21:10] isaacs: digitalspaghetti: i have no idea how much no.de instances will cost.  but yes, there will definitely be some kind of hobbyist-friendly small/cheap/free-ish deal.  probably not free, or at least, not indefinitely so, but cheap.  no.de as a product won't succeed without that, imo.
[21:11] isaacs: digitalspaghetti: it's also important to have giant expensive packages, and an easy way to move from small to huge
[21:12] creationix_: bradleymeck: where can I get plain gcc for osx?
[21:13] bradleymeck: most modern macs have it preinstalled i thought, sec
[21:15] creationix_: -bash: gcc: command not found
[21:15] creationix_: nope
[21:16] bradleymeck: well seems i was wrong, the gcc install from macports requires xcode to compile
[21:16] tjholowaychuk: cant you just curl it?
[21:17] tjholowaychuk: wait
[21:17] tjholowaychuk: duh
[21:17] tjholowaychuk: haha
[21:17] sonnym has joined the channel
[21:18] creationix_: hah, I found a snow-leopard dvd
[21:18] creationix_: not the newest xcode, but good enough
[21:18] Me1000 has joined the channel
[21:18] isaacs: creationix: so, to create a standalone gcc for mac, you can do this:  1. install Xcode 2. install macports 3. sudo port dmg g++ (probably mixed up that command a bit)
[21:19] isaacs: creationix: then that should create a standalone dmg file with just that one port
[21:19] nroot7 has joined the channel
[21:19] creationix_: ...
[21:19] isaacs: but i don't really know if that works, because i always get past step 1, and the motivation is gone
[21:19] Nohryb has joined the channel
[21:19] Gruni has joined the channel
[21:20] tjholowaychuk: bradleymeck: used SetIndexedPropertyHandler() at all?
[21:20] creationix_: yeah, once you have xcode, it works
[21:20] bradleymeck: yes
[21:20] itissid: creationix: I installed wheat with npm install wheat... DOes that solve the dependency issues...?
[21:20] creationix_: itissid: it should
[21:21] bradleymeck: tjholowaychuk whats the q?
[21:21] mjr_: Another way to avoid this problem is to always install XCode when you install a new Mac.  Just consider it part of the OS.
[21:21] creationix_: itissid: I'd npm install spark too if you want to use spark to run the app
[21:21] creationix_: but that's optional
[21:21] tjholowaychuk: bradleymeck: foo[i] = 'bar'; any reason my setter would never get called?
[21:21] creationix_: mjr_: true, but it's not my laptop
[21:21] tjholowaychuk: the getter works great
[21:21] margle has joined the channel
[21:21] bradleymeck: i is not a number
[21:22] creationix_: as soon as I'm done I'm going to add node v0.2.3 to ivy-bin
[21:22] isaacs: creationix: add the more recent patch release, plz
[21:22] creationix_: isaacs: what do you mean?
[21:22] tjholowaychuk: bradleymeck: well even foo[0] = whatever
[21:22] tjholowaychuk: wont work either
[21:22] isaacs: creationix: 0.2.3-3
[21:23] isaacs: http://registry.npmjs.org/npm/0.2.3-3
[21:23] bradleymeck: mmm, can you use SetNamedPropertyHandler?
[21:23] creationix_: isaacs: npm version v0.2.3-3?
[21:23] margle: I'm using querystring to parse an HTTP POST, on the html side, how would I format/ name the input elements to get an array of objects?
[21:23] creationix_: I was talking about node versions
[21:23] margle: [ {}, {}, {} ]
[21:23] isaacs: creationix: yeah
[21:24] creationix_: on my mbp "Install time remaining: About 34 minutes" What the heck is it doing?
[21:24] boaz_ has joined the channel
[21:24] tjholowaychuk: bradleymeck: will try. I could probably use the pixel array stuff, but cairo orders grba instead of rgba so not sure if that would work (barely touched the v8 api)
[21:24] bradleymeck: smexing up your kernel?
[21:25] bradleymeck: i would avoid pixel array until it gets more fleshed out to api consumption
[21:30] robotarmy has joined the channel
[21:31] dmcquay has joined the channel
[21:31] itissid: creationix: So all i need to do is run wheat is "node wheat.js  myBlogRepo" right? 
[21:31] creationix_: itissid: so wheat is a library
[21:31] creationix_: itissid: http://github.com/creationix/howtonode.org/blob/master//app.js
[21:32] creationix_: that's how I use it for howtonode.org
[21:32] creationix_: wheat is a connect middleware layer
[21:32] itissid: creationix: Aha!
[21:32] creationix_: actually, that's not up to date, let me get the real script 
[21:33] creationix_: itissid: here is the real script runnong howtonode.org right now https://gist.github.com/fd0b22ac22fc0f673f84
[21:34] creationix_: disabled gzip because it was unstable and added pretty stack traces for errors
[21:34] itissid: ok so just "node app.js" and let it find wheat... Hmm let me try
[21:34] creationix_: no, it's packaged as a spark app currently
[21:34] creationix_: notice how I export the server as module.exports
[21:34] creationix_: just type spark in the same directory as app.js
[21:35] creationix_: (once you've installed spark from npm)
[21:36] creationix_: itissid: if you don't want to use spark, then just call .listen on the server object instead of exporting it.  Then you can launch it with "node"
[21:37] dysinger has joined the channel
[21:38] stephenjudkins has joined the channel
[21:39] ezmobius has joined the channel
[21:39] prettyrobots has joined the channel
[21:40] robotarm_ has joined the channel
[21:42] creationix_: Install homebrew, install xcode, install git and wget (via brew) install nvm, install node from nvm, install npm, npm install wheat spark
[21:42] creationix_: wow, that's a lot of steps
[21:42] isaacs: creationix: yep.
[21:42] tjholowaychuk: if you used ndistro it would be one command :D
[21:42] isaacs: tjholowaychuk: only if you already had wget
[21:42] creationix_: ndistro installs xcode?
[21:42] isaacs: tjholowaychuk: and gcc
[21:42] tjholowaychuk: isaacs: or curl
[21:43] tjholowaychuk: isaacs: nope precompiled
[21:43] tjholowaychuk: only
[21:43] tjholowaychuk: and no git
[21:43] isaacs: tjholowaychuk: ew.
[21:43] tjholowaychuk: its great
[21:43] lianj: haha, maybe make a locomotiv thingy like rails did for its mac users that just wanted to try node/rails
[21:43] isaacs: tjholowaychuk, creationix: srsly, why not just package it up into a .app?
[21:43] creationix_: I think I'm teaching a node intro course this weekend in Santa Clara, I made provide something like ndistro to my students
[21:44] creationix_: s/made/might/
[21:44] isaacs: tjholowaychuk, creationix: put the .app in a .dmg, with a picture that shows you dragging it to a shortcut to /Applications.  then run "Node.app"
[21:44] creationix_: lol
[21:44] creationix_: isaacs: now we're talking
[21:44] creationix_: how would that Node.app provide command-line utils?
[21:45] hassox has joined the channel
[21:45] isaacs: write some dumb gui in Applescript, give it a File > Open... and select a js file, and it does osascript -e 'tell application Terminal to open and enter text "path/to/node file.js"'
[21:45] isaacs: creationix: through menu options
[21:45] creationix_: interesting idea
[21:45] isaacs: imo, there are three deployment scenarios one should care about.  the server, the client, and the gui
[21:45] creationix_: a simple node ide for dummies then?
[21:45] isaacs: don't try to make a client app as easy as a gui app.  it will fail.  just make a gui app.
[21:45] isaacs: it's not terrifically hard.
[21:46] derencius has joined the channel
[21:46] isaacs: take a look at CouchDBX.  just a browser window pointing at futon, and a window that tails the log, and a start/stop button.
[21:46] creationix_: true
[21:46] isaacs: everything bundled in the .app
[21:47] isaacs: stores its stuff in ~/Library/Application Support/CouchDBX
[21:47] creationix_: but node basically provides "node" and "npm" and maybe "spark" as command-line tools
[21:47] isaacs: it's great
[21:47] creationix_: maybe a way to add the bin folder in the app to the user's path?
[21:47] creationix_: that might work
[21:47] creationix_: and the gui could have a basic interface to npm 
[21:47] isaacs: creationix: no, you put *everything* in the .app folder, and give them a menu option to link the bins into /usr/local/bin or some such.  like TextMate does
[21:48] isaacs: creationix: but really, i think node is a bad fit for that.
[21:48] creationix_: path/link sure, same effect
[21:48] isaacs: creationix: look at the awfulness that is mamp
[21:48] creationix_: yeah, I don't want to do too much
[21:48] isaacs: if it's a gui app, it is a failure if you *ever* have to drop down to the cli
[21:48] isaacs: tha'ts a MAY, never a MUST
[21:49] creationix_: maybe an installer that installs node and npm into /usr/local would be better?
[21:49] isaacs: or into ~/local
[21:49] creationix_: ~/local/bin isn't
[21:49] isaacs: creationix: or just tell people to get a compiler.
[21:50] isaacs: creationix: btw, the "npm sucks as root" thing will be much more pretty soon.
[21:50] creationix_: well, if they come to my 75 minute class without a compiler, then they're out of luck
[21:50] isaacs: if not the next version, then the one after that
[21:50] itissid: creationix: ok installed spark... http://itissid.com:3000/ Weird...
[21:50] creationix_: itissid: pretty
[21:50] itissid: Something is really fucked up...
[21:50] tjholowaychuk: creationix_: ndistro! lol just throw a 0.2.3 in there, then they wont need git or anything
[21:50] bradleymeck: is there a way to forcibly exit a spawned child?
[21:51] tjholowaychuk: and you can explain require paths since shit is available in ./libs/node
[21:51] tjholowaychuk: s/libs/lib/
[21:51] creationix_: itissid: what node version do you have?
[21:52] itissid: v0.3.0-pre
[21:52] creationix_: I don't think it works with v0.3.x yet
[21:52] itissid: Yikes!
[21:52] creationix_: looks like the buffer api changed
[21:53] skepticsean has joined the channel
[21:53] creationix_: I wonder where all the binary-looking stuff is coming from though
[21:53] seanouimet has joined the channel
[21:55] sprout has joined the channel
[21:56] emmanueloga has joined the channel
[21:56] itissid: every rtime i reload its someting funny and different
[21:56] itissid: :)
[21:57] evanpro has joined the channel
[21:58] franksvalli1 has joined the channel
[21:58] creationix_: itissid: yeah, it's failing on the first file read, but wheat loads several files in parallel
[21:58] Nohryb has joined the channel
[21:58] creationix_: so it's a race condition to see which errors first
[21:58] saikat has joined the channel
[21:59] cloudhead has joined the channel
[22:00] itissid: Creationix any version <0.3.x works right?
[22:01] itissid: I will use nvm :)
[22:01] creationix_: yeah, about v0.1.104-v0.2.3
[22:02] charlenopires has joined the channel
[22:03] secoif has joined the channel
[22:03] creationix_: hmm, I thought openssl should be found on osx
[22:03] creationix_: I even installed pkg-config
[22:03] creationix_: still says Checking for openssl                     : not found 
[22:04] itissid: creationix: i think i got that yesterday
[22:04] itissid: openssl is in  the path?
[22:04] creationix_: well, I also got Checking for header openssl/crypto.h     : yes 
[22:04] itissid: install openssldev ..
[22:04] itissid: thats how i solved
[22:04] creationix_: so it must be there, at least partially
[22:04] tilgovi: can I just set exports = function (... ) {... }
[22:04] creationix_: openssldev? on osx?
[22:04] tilgovi: and then the return from require is callable?
[22:05] creationix_: tilgovi: module.exports = function (…) {…}
[22:05] tilgovi: thanks
[22:05] tilgovi: ahh. that's because I don't want to rebind the symbol "exports"
[22:05] tilgovi: exports.foo = ... works fine, but to change the module's exports object entirely, I need to go above it
[22:06] creationix_: hmm, HEAD can find openssl
[22:06] itissid: well i dunno the name osX... i think its called libssldev
[22:06] creationix_: tilgovi: yep
[22:06] itissid: I was doing a node install on my chunkhost debian system when i got the error..
[22:07] nrub` has joined the channel
[22:07] technoweenie has joined the channel
[22:08] kjy112 has joined the channel
[22:09] vnguyen has joined the channel
[22:09] creationix_: yeah on linux it's libssl-dev
[22:09] creationix_: but on osx with xcode the headers are already there
[22:10] creationix_: s/linux/debian\/ubuntu/
[22:10] itissid: creationix: I did a node down grade to 0.2.3  and ran wheat again... still no love.. http://itissid.com:3000/
[22:10] digitalspaghetti: anyone done anything with nodejs and gtk?
[22:10] digitalspaghetti: or any other windowing library?
[22:10] itissid: I used nvm..
[22:10] creationix_: itissid: and you reinstalled npm for the new node?
[22:11] itissid: damn.. One sec...sorry
[22:11] rex_fernando has joined the channel
[22:11] creationix_: type `which spark` to see which version you're using
[22:11] creationix_: nvm doesn't change any global links so other terminals aren't affected
[22:12] bradleymeck: digitalspaghetti i started on sfml which is minimal but work is killing me from now til thanksgiving :(
[22:13] itissid: creationix: spark@0.1.2
[22:13] creationix_: itissid: sure, what which spark binary
[22:13] creationix_: nvm has a different set of libraries for each node version
[22:14] creationix_: for example, my output is "/Users/tim/.nvm/v0.2.3/bin/spark"
[22:14] creationix_: s/binary/executable/
[22:17] itissid:  ok so /home/sid/.nvm/v0.2.3/bin does not have any spark... Although i just did reinstalled npm(curl http://npmjs.org/install.sh | sh) and the spark from: npm install spark 
[22:17] creationix_: did you do both from the same bash session?
[22:17] jakehow has joined the channel
[22:17] creationix_: you need v0.2.3 activate in the session that installs npm and spark
[22:18] creationix_: *active
[22:18] itissid: Ok its spark v0.1.2
[22:18] paul__ has joined the channel
[22:18] aheckmann has joined the channel
[22:19] itissid: now its all done in one bash shell
[22:19] creationix_: so does "which spark" give /home/sid/.nvm/v0.2.3/bin/spark?
[22:19] itissid: Hmm
[22:19] itissid: yeah it does
[22:20] digitalspaghetti: interesting: http://blog.jquery.com/2010/10/04/new-official-jquery-plugins-provide-templating-data-linking-and-globalization/
[22:20] digitalspaghetti: i wonder how easy it would be to port that templating library over
[22:21] itissid: We have love.. :) http://itissid.com:3000/
[22:21] creationix_: itissid: looks like it's working now
[22:21] creationix_: :)
[22:21] creationix_: now to fix wheat to not puke on v0.3.x stuff
[22:21] creationix_: if possible
[22:21] loincloth has joined the channel
[22:21] itissid: Thanks dude... 
[22:21] creationix_: no problem
[22:22] MikhX has joined the channel
[22:22] dguttman has joined the channel
[22:23] itissid: creationix: yeah its a real crazy dance if you are on version 0.3.x and you have to dwngrade... i hope you can get things done on 0.3.x.. god luck..
[22:26] creationix_: isaacs: applescript looks weird
[22:26] creationix_: is there a way to use javascript instead?
[22:26] isaacs: creationix: it does
[22:26] halfhalo: applescript is like retarted javascript
[22:26] isaacs: creationix: yes, but not really
[22:26] sschuermann:  npm update killed my express ap 
[22:26] derferman has joined the channel
[22:26] itissid: Did some one check out chunkhost.. They are offereing free machines, no wait time... grab em while you can...!
[22:26] sschuermann: will file a bug 2 morrow .. looks like a jade problem 
[22:27] isaacs: sschuermann: oh?
[22:27] isaacs: ok
[22:27] sschuermann: oh? 
[22:27] rwhitby has left the channel
[22:27] tjholowaychuk: sschuermann: i tried pushing a new jade release this morning while npm was semi-busted
[22:27] tjholowaychuk: sschuermann: whats the issue?
[22:27] sschuermann: ah ok 
[22:27] sschuermann: node.js:275
[22:27] sschuermann:         throw new Error("Cannot find module '" + request + "'");
[22:27] sschuermann:               ^
[22:27] sschuermann: Error: Cannot find module './../.npm/jade/0.5.1/package/lib/jade/index'
[22:27] sschuermann:     at loadModule (node.js:275:15)
[22:27] sschuermann: sorry .... 
[22:28] sschuermann: ill recheck 2 morrow and file a bug rep if still there 
[22:28] tjholowaychuk: ohhhh
[22:28] tjholowaychuk: I know what thaht is shit
[22:28] tjholowaychuk: changed the structure a bit but didnt update package.json
[22:29] sschuermann: i should put all my stuff under ci to monitor you guys ;) 
[22:29] sschuermann: this is one of the cases i wanted to write tests for the examples ;) 
[22:29] tjholowaychuk: sschuermann: should be ok now
[22:29] sschuermann: ty
[22:29] bpadalino has joined the channel
[22:30] sschuermann: Error: ENOENT, No such file or directory '/usr/local/lib/node/.npm/express-1.0.0rc2/package/package.json'
[22:30] vnguyen has joined the channel
[22:30] sschuermann: on npm update 
[22:30] sschuermann: no need to stress out 
[22:30] isaacs: sschuermann: a bug that is fixed in the latest release.  npm update npm, and then it should work
[22:31] isaacs: sschuermann: if "npm update npm" says "nothing to update", then you canjust "npm install express" to get the latest version, or "npm rm express" to remove all trace of it.
[22:31] isaacs: sschuermann: "npm update" is basically just sugar around other commands.
[22:31] sschuermann: npm update didnt work 
[22:31] sschuermann: 
ACTION must follow the sudont advice 

[22:31] isaacs: sschuermann: not "npm update".  you need to do "npm update npm"
[22:31] sschuermann: i didt told me i am up2date
[22:32] isaacs: ok, cool
[22:32] isaacs: bbiab, meeting
[22:32] csullivan has joined the channel
[22:32] sschuermann: ah rm and install 
[22:33] sschuermann: hmm .. ill file a bug 2morrow 
[22:35] c4milo has joined the channel
[22:37] Tim_Smart has joined the channel
[22:38] rex_fernando has joined the channel
[22:39] creationix_: isaacs: I think one of our QT guys is going to help me make a super minimal node ide for osx + windows
[22:39] creationix_: if only cygwin wasn't needed for windows
[22:40] Tim_Smart: Pfft IDE. vim + terminal <3
[22:40] creationix_: well, for getting started, a simple native app with pre-built node binaries can't be beat
[22:41] creationix_: Tim_Smart: btw, vim is an ide
[22:41] creationix_: ;)
[22:41] c4milo: anybody knows a xml2json library in C or cpp ? 
[22:41] stephenjudkins has joined the channel
[22:42] creationix_: c4milo: the problem is that they don't match very well
[22:42] c4milo: creationix_:  why ? 
[22:42] creationix_: because xml is much richer in structure than json
[22:42] creationix_: json is simply nested key/value with a few prmitives
[22:43] creationix_: <a><b><b><a> -> json ?
[22:43] c4milo: creationix_: hum you mean that the existing libraries doesn't work well ? 
[22:43] Tim_Smart: { a: { b: ... } }
[22:43] creationix_: c4milo: no, that xml isn't the same abstract structure as json
[22:44] creationix_: it's impossible to write a generic converter that works for all cases and round-trips to the same thing
[22:44] c4milo: so what is your advice ? 
[22:44] creationix_: Tim_Smart: sorry, I meant <a><b/><b/></a>
[22:45] creationix_: c4milo: find an xml parser and either using sax or dom write your own converter to the json you want
[22:45] creationix_: but it's going to require some domain knowledge
[22:45] tjholowaychuk: it would not be pretty, but you could easily do it
[22:45] tjholowaychuk: like Tim_Smart said, except a lot more arrays
[22:45] CrabDude has joined the channel
[22:46] creationix_: tjholowaychuk: right, you can preserve the xml structure by generating super verbose json
[22:46] creationix_: but that's almost never what you want
[22:46] mbrochh has joined the channel
[22:46] mbrochh has joined the channel
[22:46] Tim_Smart: creationix: Considering DOM is practically JSON: { document: { textContent: ..., childNodes: ... } }
[22:46] creationix_: and if you converted it back to xml, it would become even more bloated and verbose
[22:47] creationix_: <person><name>Tim</name><age>28</age></person> -> {"
[22:47] pufuwozu has joined the channel
[22:47] creationix_: {"name":"Tim","age":28}
[22:47] creationix_: but that only works for a subset of xml
[22:48] creationix_: make that {"person":{"name":"Tim","age":28}}
[22:48] bradleymeck: ultra verbose json can be the same as xml, but it is extremely hard to work with from what i find for any use
[22:48] creationix_: yep
[22:48] derferman has joined the channel
[22:48] galaxywatcher has joined the channel
[22:48] bradleymeck: you end up with node[0].x[1].y[2] etc
[22:49] creationix_: [{"tag":"person","children":[{"tag":"name","children":["Person"]},{"tag":"age","children":["28"]]}]
[22:49] bradleymeck: i think there is a jsonml converter around somewhere if that is what you want though
[22:49] creationix_: something like that right?
[22:50] dpritchett has joined the channel
[22:50] creationix_: I guess you could use a standard jsonml converter and then post-process the json to simplify it down using your domain knowledge
[22:52] inimino: creationix_ ⋱ It's not quite that verbose.
[22:53] tjholowaychuk: creationix_: you can make all those objects array-like
[22:53] tjholowaychuk: or arrays with props
[22:53] inimino: <x a="b"><y/></x> -> ["x",{"a":"b"},["y"]]
[22:53] tjholowaychuk: to knock off .children
[22:53] tjholowaychuk: might print funky though
[22:53] creationix_: inimino: that's looks like my haml-js
[22:53] inimino: creationix_ ⋱ That's JSONML IIRC.
[22:54] creationix_: oh right, I remember seeing that
[22:54] inimino: An element becomes an array, first element is tag name, second is attribute dictionary, third is children.
[22:54] creationix_: but that's for json -> xml
[22:54] ginader has joined the channel
[22:54] creationix_: well, I guess it can work the other way
[22:55] techwraith has joined the channel
[22:55] inimino: yeah, it's intended to
[22:55] creationix_: http://github.com/creationix/jquery-haml
[22:55] inimino: I still think it's verbose, but it's not much more verbose than XML itself.
[22:56] techwraith: Does anyone know if there's a Node RSS parser that can also create RSS feeds?
[22:56] creationix_: techwraith: no, but it sounds interesting
[22:56] creationix_: I'm using my haml-js to generate rss for howtonode.org
[22:56] creationix_: it's painful
[22:56] hober: It'd be nice to port the UFP to node
[22:57] tjholowaychuk: creationix_: why is it painful?
[22:57] tjholowaychuk: i remember doing that lots with ruby haml
[22:57] creationix_: as opposed to a higher abstraction
[22:57] techwraith: Yeah, that's kind of what I was hoping for - didn't want to reinvent the wheel if it was out there already.
[22:58] creationix_: something that's rss aware and you just input the data, not the xml
[22:59] techwraith: Is howtonode.org open source?
[23:04] unomi: techwraith: the articles or the app?
[23:04] unomi: http://github.com/creationix/howtonode.org
[23:04] techwraith: The app, found it already, thanks :)
[23:04] techwraith: Wheat looks cool.
[23:04] unomi: Site Design and code is © 2010 to Tim Caswell under the MIT license. Content and articles are copyrighted to the individual authors. All code snippets used in the examples are in the public domain.
[23:04] creationix_: techwraith: thanks
[23:05] pgriess has joined the channel
[23:05] unomi: creationix_: the site looks a bit odd in my chrome at the moment
[23:05] creationix_: unomi: how so?
[23:05] pgriess has joined the channel
[23:05] ajpiano has joined the channel
[23:05] unomi: the right hand sidebar is under the content
[23:05] creationix_: unomi: cache issue, do a few hard refreshes
[23:05] creationix_: your css is old, the the html is new
[23:06] unomi: there we go
[23:06] creationix_: I should have renamed the css file to prevent that, oh well
[23:06] bmizerany has joined the channel
[23:06] zpoley has joined the channel
[23:07] unomi: or attached a querystring
[23:08] unomi: even if the underlying name is the same site.css != site.css?v=1
[23:08] creationix_: that works too
[23:08] unomi: and you can control that from the html alone
[23:08] jashkenas has joined the channel
[23:09] creationix_ has joined the channel
[23:10] rex_fernando: so why are there like 20 different web frameworks for node?
[23:10] bruse: because like 20 different people wanted to create their own
[23:10] konobi: why not?
[23:10] unomi: rex_fernando: frameworks are the new CMS
[23:11] micheil: techwraith: oh, hey.
[23:12] techwraith: micheil: Oh, hey. :)
[23:12] creationix_: unomi: ok, added query string
[23:12] rex_fernando: so are any of them community favorites?
[23:13] unomi: rex_fernando: try a few and see what works
[23:13] micheil: rex_fernando: I'd guess at express, connect, geddy, and one of the others being the community favourites
[23:13] unomi: I seem to be seeing more buzz around express than others though
[23:13] micheil: I personally like express in many ways, but could probably use each of the others if I wanted to
[23:13] creationix_: btw, connect isn't a web framework, it's a framework's framework
[23:13] creationix_: you can build your own on top of connect
[23:13] creationix_: like express did
[23:14] micheil: creationix: same diff, really.
[23:14] creationix_: except it's not designed to be easy to the end-developer
[23:14] unomi: btw, has anyone come across express doing html output prettyfying?
[23:14] creationix_: and is super modular
[23:14] lianj: rex_fernando: choose one that uses like the js version of rack. thats should be the common framework on with web frameworks build on and share quality
[23:14] tjholowaychuk: unomi: that is up to the template engine
[23:14] creationix_: lianj: that's what connect aims to do
[23:14] atmos has joined the channel
[23:14] rex_fernando: is there any good reason to build your own?
[23:14] creationix_: also there is jsgi
[23:14] lianj: creationix_: just sucks at PR
[23:14] unomi: tjholowaychuk: Thanks
[23:15] unomi: I have been playing around with the YUI3 + Express stack Dav Glass put up
[23:15] tjholowaychuk: unomi: but since webkit (and others) display the dom in a nice way most ppl dont bother
[23:15] rex_fernando: I tried express, seemed nice. But then I realized there were a bunch others
[23:15] creationix_: rex_fernando: yes, frameworks are a royal pain if they don't fit your use case
[23:15] unomi: its pretty funky
[23:16] creationix_: things like rack, connect, and jsgi allow you to not have to reinvent the whole wheel to make a custom system
[23:16] tjholowaychuk: unomi: yeah I would not recommend fiddling with a dom for your template engine lol
[23:16] franksvalli has joined the channel
[23:16] charlenopires has joined the channel
[23:16] unomi: tjholowaychuk: well, its more for building javascript widgets serverside
[23:16] creationix_: someone could write a html pretty printer as a connect middleware
[23:16] unomi: the templating aspect is just a bonus, if you want. 
[23:17] rex_fernando: ok cool, thanks
[23:17] jameshome_: is anyone working on html sanitization connect middleware?
[23:17] digitalspaghetti: if we can get ExtJS working that way, it would be nice to build widgets on the fly and load them as components
[23:17] unomi: I generally try not to output html at all
[23:18] digitalspaghetti: unomi: yea, i'm using Ext Direct which is RPC calls, so i'm all JSON
[23:18] unomi: but its pretty cool to be able to offer a functional version of your site to people who don't have js enabled
[23:18] unomi: or as a way to get a fast initial load 
[23:18] digitalspaghetti: i'm doing an app not a website, so you will need Firefox or Chrome
[23:19] digitalspaghetti: hopfully IE9 as well, i'm not using any socket stuff
[23:19] creationix_: digitalspaghetti: still, the initial load is generally slow with "apps"
[23:20] unomi: I see it as slightly gimmicky, the building of widgets serverside, but what we do is all about the gimmicks anyway :p
[23:20] unomi: creationix_: right, thats the upside to serverside rendering
[23:20] unomi: you can build the html version of the js widget and set it down the wire
[23:21] unomi: then load the js under the hood while the user is making his mind up what to do
[23:21] creationix_: unomi: sounds like you've been talking to dav glass
[23:21] unomi: the next time the user visits, the js is cached, pure json is sent and checks are in the mail
[23:22] unomi: creationix_: heh, yeah, I sent him a message some months ago.
[23:22] creationix_: I think it's an interesting idea.  But until we have a faster dom for node, it's very expensive
[23:23] unomi: sure, but its only expensive for the initial load of unique clients, which is a bit of a win over traditional 'rails / php'
[23:23] unomi: while reducing the 2 - 3 stage load of traditional js apps
[23:24] galaxywatcher_ has joined the channel
[23:25] franksvalli has joined the channel
[23:27] mikeal has joined the channel
[23:27] Wandrewvious has joined the channel
[23:28] dgathright has joined the channel
[23:29] aconbere has joined the channel
[23:29] dilvie has joined the channel
[23:33] galaxywatcher has joined the channel
[23:33] jamescarr has joined the channel
[23:33] jamescarr: hmph
[23:33] stephenjudkins has joined the channel
[23:34] sprout has joined the channel
[23:34] `damien has joined the channel
[23:37] probablycorey has joined the channel
[23:39] jamescarr: isaacs, you here?
[23:39] isaacs: yo
[23:39] csullivan_ has joined the channel
[23:39] isaacs: whatssup?
[23:39] jamescarr: I have been having an odd issue with an installed module I develop on...
[23:39] isaacs: ok
[23:39] jamescarr: my tests see the installed module, not the source
[23:40] jamescarr: even though I defined the require path to the source dir... any idea why that happens? :)
[23:40] isaacs: not sure what that means
[23:40] isaacs: are you running your tests that are installed via npm?
[23:40] jamescarr: well, it's seeing the code from the version isntalled via npm
[23:40] isaacs: and this is a dependency?
[23:40] jamescarr: yeah
[23:40] jamescarr: it's a dep in another module I have installed
[23:40] isaacs: and your program is installed via npm as well?
[23:40] jamescarr: yeah
[23:40] isaacs: ok, then this is by design.
[23:41] isaacs: npm tries very hard to make *sure* that your code gets the version that you told it you want.
[23:41] isaacs: to put something else in that spot, do this:
[23:41] isaacs: 1. go to the source of the thing that you want to use
[23:41] isaacs: 2. npm link it
[23:41] jamescarr: ah
[23:42] isaacs: 3. if that doesn't update the dependents, then you could do "npm update-dependents thing@stupid-long-link-version" to try to point the deps at the link version
[23:42] nerdEd has joined the channel
[23:42] isaacs: update-dependents is a plumbing command that attempts to point dependent packages at the version you specify
[23:42] xAFFEE has joined the channel
[23:42] jamescarr: cool
[23:42] jamescarr: thanks, I'll give it a try
[23:43] isaacs: np
[23:43] isaacs: note that npm actually generates code so that this happens at run time.  it's not ideal to do things this way, but it is more flexible and reliable than environ setting.
[23:43] abiraja has joined the channel
[23:46] tmpvar has joined the channel
[23:47] xAFFEE has left the channel
[23:48] [[zz]] has joined the channel
[23:49] ekidd has joined the channel
[23:49] stephenjudkins has joined the channel
[23:51] polotek has joined the channel
[23:54] Tim_Smart: Grr.. Why can't mercurial give some sort of progress update when cloning a repo.
[23:55] Tim_Smart: I had to check my network activity to see if was hanging or not.
[23:55] halfhalo: Because... It's crap?
[23:55] polotek: problem #37 with software
[23:55] polotek: lack of secondary usability features
[23:56] tilgovi: does pump close the writeable stream if the readable stream closes?
[23:56] polotek: tilgovi: yes
[23:56] mikeal: no
[23:56] halfhalo: We talking about windows again, right?
[23:56] tilgovi: mikeal: I should listen for 'end' myself?
[23:56] mikeal: only on error
[23:56] tilgovi: okay.
[23:56] polotek: mikeal: since when?
[23:56] mikeal: since never
[23:56] mikeal: :)
[23:56] mikeal: it does it the other way
[23:56] polotek: Whatchootalkinboutwillis
[23:56] mikeal: wait
[23:57] mikeal: it closes readable on the writeable end
[23:57] mikeal: er close
[23:57] mikeal: on writable close
[23:57] mikeal: which might not be what you want
[23:57] tilgovi: nah, I want the opposite.
[23:57] mikeal: yeah, it calls end
[23:57] polotek: oh, nevermind
[23:57] chrischr1s has joined the channel
[23:57] polotek: end != close
[23:57] tilgovi: so. read until end, close writeable
[23:57] tilgovi: is what I want
[23:57] mikeal: so end event on readable calls end on writable
[23:58] polotek: it does end() the writable when the readable ends
[23:58] tilgovi: oh, nice.
[23:58] mikeal: but it doesn't call close
[23:58] mikeal: close is something else
[23:58] tilgovi: but then I should call close?
[23:58] mikeal: depends
[23:58] tilgovi: or, I should call close if it's appropriate :-P
[23:58] mikeal: what type of stream?
[23:58] tilgovi: http... I know the answer here
[23:58] mjr_: It's quite enlightening to just copy and paste sys.pump's contents into your program and fiddle with the event handlers, add logging, etc.
[23:58] mikeal: HTTPResponse doesn't have a close 
[23:58] dgathright has joined the channel
[23:59] mikeal: wait, maybe it does
[23:59] mikeal: but you don't wanna call it
[23:59] dpritchett has joined the channel
[23:59] Tim_Smart: mjr_: streams <3
[23:59] mikeal: you wanna let the parser and the client to decide
[23:59] tilgovi: oh, right.
[23:59] dpritchett: how would I go about building a socket.io client app in node.js?
[23:59] Tim_Smart: We definitely need some sort of stream template that everything derives from though.
[23:59] dpritchett: like say i built a chat room using node and socket.io and now i want to build a chat bot as a standalone program that connects
[23:59] mikeal: that's what the filter interface is for
[23:59] mikeal: i need to start that conversation back up
[23:59] mjr_: Tim_Smart: yes, unifying the stream interface will be very useful.