[00:00] mikeal: did you deploy it?
[00:00] ashb: no :)
[00:00] ashb: but its fixed which is all that counts? right >_>
[00:00] ashb: mjijackson: if you notice any problems, let me know
[00:00] ashb: i've sorted out all the edge cases i've run across, but entirely possible there are more
[00:01] derferman has joined the channel
[00:01] ashb: mikeal: thanks - guess i better add proper accept parsing in
[00:03] ashb: mikeal: hmmm so whats the priority of that last application/json?
[00:03] ashb: since its type,type;q=prio isn't it?
[00:03] ashb: (i.e. the q= applies to the stuff on the left?
[00:04] mikeal: i don't pay attention to any of that and just do left to right priority
[00:04] ashb: i thought order was irrelevant
[00:04] mikeal: it's not
[00:05] mikeal: it's left to right
[00:05] ashb: thats not what the spec says
[00:05] ashb: http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html
[00:05] ashb: Accept: audio/*; q=0.2, audio/basic SHOULD be interpreted as "I prefer audio/basic, but send me any audio type if it is the best available after an 80% mark-down in quality."
[00:05] joshholt has left the channel
[00:06] joshholt has joined the channel
[00:06] joshholt has left the channel
[00:08] mikeal: what is the default level
[00:08] ashb: q=1
[00:09] mikeal: man this is insane
[00:09] ashb: yeah it is
[00:09] mikeal: dumbest header ever
[00:09] ashb: ah so *each* mime type needs a ;q= if it shouldn't be 1
[00:09] mikeal: ok, so JSON Viewer plugin for firefox ads this header
[00:09] ashb: but left to right does apply
[00:10] mikeal: and it just tacks it on to the end :)
[00:10] ashb: hmmm tho i think i recall reading that Left->Right doesn't matter
[00:10] ashb: but it practice it probably does
[00:10] mikeal: and since firefox sets the types it actually cares about to 0.9 application/json wins
[00:10] tilgovi has joined the channel
[00:10] mikeal: this is why I ignore the accept header and just namespace the different interfaces
[00:11] ashb: the 0.9 only applies to app/xml
[00:11] ashb: i think
[00:11] ashb: text/html,application/xhtml+xml and application/json all default to ;q=1
[00:11] mikeal: who came up with this!
[00:11] ashb: a committe ;)
[00:11] mikeal: so, when you do one,two,three;p=.8
[00:12] mikeal: that only applies to that last content type?
[00:12] mikeal: so one and two are still 1?
[00:12] ashb: i think so
[00:12] ashb: according to the letter of the spec
[00:13] mikeal: remind me never to use this in any app that i ever write
[00:13] ashb: it seemed like a good idea at the time >_<
[00:13] ashb: i'm begging to regret it
[00:13] mikeal: i usually namespace the interfaces anyway and just make sure that the header exists and that the content-type i'm going to use is in there
[00:14] mikeal:  /api/json/blah/blah
[00:14] mikeal:  /api/xml/blah/blah
[00:14] mikeal: except i usually don't get around to writing the xml one :)
[00:14] ashb: i was going for something more explicitly RESTful
[00:14] mikeal: REST is great until it's confusing
[00:14] ashb: such as when browers come into the picture usually :)
[00:18] inimino: yes
[00:18] inimino: erm, I was scrolled up
[00:19] inimino: that was in reply to "p=.8 applies to the last one only"
[00:20] cloudhead has joined the channel
[00:21] ashb: hold on i think the spec contradicts itself \o/
[00:21] ashb: "Each media-range MAY be followed by one or more accept-params, beginning with the "q" parameter for indicating a relative quality factor. "
[00:21] ashb: it then goes on to give an example of "text/html;level=2;q=0.4"
[00:22] ashb: which doesn't begin with the "q" parameter
[00:23] inimino: hm...
[00:24] inimino: hm, that's interesting
[00:24] ashb: it is, isn't it?
[00:26] inimino: oh, I see
[00:26] ashb: did i read it wrong?
[00:26] inimino: level=1 is a parameter
[00:27] inimino: it's part of the media-range production
[00:27] inimino: which comes before the accept-params
[00:27] ashb: oh;
[00:27] ashb: i missed that
[00:27] ashb: where is the parameter production?
[00:27] inimino: that's why they have that note about how you can't have media type parameters named 'q'
[00:28] inimino: I don't see it, it's probably aliased to 'token' or something somewhere later
[00:28] ashb: yeah param = attr "=" value
[00:28] ashb: attr = token
[00:28] ashb: vlaue = token | quoted-string
[00:28] softdrink has joined the channel
[00:28] ashb: (§3.6)
[00:29] ashb: eh sod it: .split(/\s*,\s*/) will do
[00:29] inimino: not one of my favorite specs
[00:29] inimino: hehe
[00:29] ashb: its technically wrong as you could have a , inside a quoted string
[00:30] ashb: but like i give a damn about that - its damn unlikely to happen
[00:30] inimino: hehe
[00:31] derferman has joined the channel
[00:32] Tim_Smart: Anyone here listened to the Ajax.org podcast from thechangelog?
[00:37] _ry: what's the best way to set a getter on a prototype?
[00:38] ashb: Object.defineProperty probably :)
[00:39] _ry: is Object.defineProperty(Class.prototype, 'foo', fn) 
[00:39] _ry: going to work? :/
[00:42] Tim_Smart: _ry: Usually the most universal way is: Object.__defineGetter__('foo', fn);
[00:42] ashb: js> Object.__defineGetter__
[00:42] gbot2: ashb: <function __defineGetter__() { [native code] }>
[00:42] Tim_Smart: where Object is your prototype etc
[00:42] Tim_Smart: so: Class.prototype.__defineGetter__('foo', fn);
[00:42] ashb: won't that define it on the Object prototype itself
[00:43] ashb: yeah
[00:43] ashb: _ry: v8 shoud get a fucking move on and implment ES5
[00:43] ashb: cos Object.defineProperty & Object.create fucking rawk.
[00:44] _ry: object.defineProperty works in v8, it seems
[00:44] ashb: including enumerable: false ?
[00:44] Tim_Smart: js> var obj = { get test() { return 5; } }; obj
[00:44] gbot2: Tim_Smart: {test:5}
[00:44] ashb: thats been working for a while
[00:44] Tim_Smart: gbot is spidermonkey I take it
[00:44] _ry: node>  var obj = { get test() { return 5; } }; obj                                            
[00:44] _ry: { test: [Getter] }
[00:45] cloudhead: the enumerable stuff works too
[00:45] ashb: js> var obj = {}; obj.defineProperty("foo", { value: 2, enumerable: false }); obj
[00:45] gbot2: ashb: Error: TypeError: obj.defineProperty is not a function 
[00:45] ashb: js> var obj = {}; obj.defineOwnProperty("foo", { value: 2, enumerable: false }); obj
[00:45] gbot2: ashb: Error: TypeError: obj.defineOwnProperty is not a function 
[00:45] ashb: er, O.
[00:45] _ry: ashb: :)
[00:45] cloudhead: yes
[00:45] ashb: js> var obj = {}; Object.defineProperty(obj, "foo", { value: 2, enumerable: false }); obj
[00:45] gbot2: ashb: Error: TypeError: Object.defineProperty is not a function 
[00:46] ashb: doesn't exist in what ever gbot is running then
[00:47] whoahbot has joined the channel
[00:47] stepheneb has joined the channel
[00:49] cloudhead: yea it must be running an older version
[00:49] ashb: it only landed recently - we had to shim it in
[00:56] ashb: http://pastie.org/862439
[00:57] ashb: i think that covers the Accept header crap
[00:57] ashb: or at least most of the common uses of it
[00:59] Tim_Smart: nodejs_v8: "My node.js probably needs to be upgraded to latest head"
[00:59] nodejs_v8: Tim_Smart: "My node.js probably needs to be upgraded to latest head"
[01:02] nodejs_v8 has joined the channel
[01:03] Tim_Smart: nodejs_v8: "Duh, forgot to re-build the evalcx addon
[01:03] nodejs_v8: Tim_Smart: Exception: SyntaxError: Unexpected token ILLEGAL
[01:03] Tim_Smart: nodejs_v8: "Duh, forgot to re-build the evalcx addon"
[01:03] nodejs_v8: Tim_Smart: "Duh, forgot to re-build the evalcx addon"
[01:03] Tim_Smart: still works, good
[01:04] dnolen has joined the channel
[01:07] nejucomo has joined the channel
[01:09] nejucomo: Is anyone working towards making node.js compliant with ECMAScript 5 Strict mode?  http://ejohn.org/blog/ecmascript-5-strict-mode-json-and-more/
[01:10] nejucomo: I am unsure if V8 supports it yet.
[01:10] Tim_Smart: nejucomo: That job belongs to the V8 team
[01:11] nejucomo: Not exactly.  Node.js would also have to conform, and that impacts many design decisions.
[01:11] _ry: from the changelong "I think there are also possibilities to simplify [node's promises], and that's what we're currently researching, if
[01:11] _ry: except that its secretly not blocking in the backend."
[01:11] _ry: oh to be young!
[01:12] aryounce has joined the channel
[01:12] Tim_Smart: _ry: Yeah that Ajax.org podcast was pretty hilarious. Sounds like they are too busy drinking beer, and can't keep up
[01:13] nejucomo: It may be possible, using strict mode, to make node.js capabilities secure, such that you could have mobile code housed in a common server framework.
[01:14] andy_l has joined the channel
[01:14] nejucomo: I'm uncertain how feasible this is, currently, but it seems like Node.js + strict mode is not too far off.
[01:14] _ry: oops i got the quote wrong
[01:14] _ry: I think there are also possibilities to simplify [node's promises], and
[01:14] _ry: that's what we're currently researching, if we can take an approach with
[01:14] _ry: server side javascript that essentially allow people to write blocking code
[01:14] _ry: top down except that its secretly not blocking in the backend.
[01:15] siculars has joined the channel
[01:17] isaacs: _ry: with a pre-processor that compiles to non-blocking JS, that's possible
[01:17] _ry: i think unless you do green processes, like erlang - there is no way to abstract non-blocking apis in a way that doesn't suck
[01:17] Tim_Smart: _ry: We could essentially do that with registerExtension
[01:17] _ry: isaacs: that's going to have to be a very smart pre-processor
[01:17] Tim_Smart: but I agree, it would suck
[01:17] isaacs: _ry: yeah, that Narrative thing that kriszyp mentioned on the list would do the trick
[01:18] isaacs: _ry: basically, instead of doing fs.readfile(file, function (foo) { ... }) you'd do foo = fs->readfile(file); and it'd wrap up everything after it in a cb
[01:18] Tim_Smart: people just need to grow up, and learn how to do async
[01:19] isaacs: Tim_Smart: yeah, i mean, coders could just man up and use the javascript we have
[01:20] isaacs: but the preprocessor doesn't have to be all THAT smart.
[01:20] _ry: i think it's going to be more of an annoyance than anything
[01:20] isaacs: it's not quite  a regexp replace, but almost.
[01:20] _ry: but waht about when you debug? what are you going to see?
[01:20] Tim_Smart: I mean abstracting it away, just introduces more element for confusion tbh
[01:20] _ry: there is so much more then just throwing the rest of the code into a callback
[01:20] isaacs: _ry: that's why i don't trust computers to write my code for me ;)
[01:21] isaacs: _ry: right, you also need to maintain line numbers (ideally), and convert return and throw into cb(null, val) and cb(er)
[01:21] isaacs: _ry: but very much, this is not something that belongs in node.  this is a userland problem.
[01:22] nejucomo: The "non-async" goal is funny, because then you have the can-of-worms of synchronization to worry about.  Why re-invent sync problems?
[01:22] isaacs: just open the door with registerExtension or something, and see what we get.  coffeescript, mu, narrative, and parenscript will likely be early wins
[01:22] isaacs: but there's a whole family of compiled-to-js languages ready to be harnessed.
[01:23] isaacs: objj, pyjs, etc.
[01:23] isaacs: or was that pyjamas?  i forget
[01:23] isaacs: oh, haXe, too
[01:24] _ry: i think the problem is, programmers think they're smart
[01:24] _ry: when actually they're just fucking javascript programmers
[01:25] _ry: we shouldn't try to be smart - it's too ahrd
[01:25] isaacs: yeah... fucking javascript prog---hey!
[01:25] _ry: leave that to the OS programmers
[01:25] isaacs: true, being too smart is a bad idea.  just look at OS programmers.
[01:26] isaacs: earlier today, i tried to send a friend a font file that i'd exported from a zip file.  Finder SWORE it was 16kb, but the actual file was some kind of magic 0 byte pointer.
[01:26] Tim_Smart: look at linus, I swear he has arrogance problems
[01:26] isaacs: osx is way too smart.
[01:26] _ry: but linus is smart he can solve the non-blocking thing :)
[01:26] _ry: we should just bind posix
[01:27] maushu has joined the channel
[01:27] Tim_Smart: linux gave us git, so he is cool
[01:27] Tim_Smart: but his talks make me lol
[01:27] isaacs: and git is awesome because it goes so far out of its way to not be too clever.
[01:27] maushu: So, anyone knows where I can find "openid for dummies"? I need to create a library for it.
[01:27] Booster has joined the channel
[01:27] stepheneb has joined the channel
[01:29] kriskowal has joined the channel
[01:31] dnolen has joined the channel
[01:37] brapse has joined the channel
[01:38] nejucomo has left the channel
[01:41] creationix has joined the channel
[01:42] rtomayko has joined the channel
[01:45] joshholt has joined the channel
[01:48] inimino: ashb: yes, gbot2 is spidermonkey
[01:49] creationix: _ry: do you have any comments or suggestions on the new doc site, or should I keep plodding along?
[01:50] pedrobelo has joined the channel
[01:50] Tim_Smart: nodejs_v8: Object.defineProperty
[01:50] nodejs_v8: Tim_Smart: function defineProperty() { [native code] }
[01:50] _ry: creationix: it's lovely
[01:51] creationix: do we want the content to be indexable by search spiders?
[01:51] creationix: because it's not currently
[01:51] creationix: of course the source on github is though
[01:52] cedricv has joined the channel
[01:52] _ry: yes, i haven't looked at how this works but the whole thing should be loaded in one html file
[01:52] creationix: I could make a hybrid version that served static content using node and rendered dynamic content in the browser for everyone else
[01:52] _ry: and just show/hide sections
[01:53] _ry: how is it working right now?
[01:53] creationix: it ajaxes in the markdown directly
[01:53] creationix: and renders the interface using client-side views
[01:53] _ry: a single file would be rad
[01:54] _ry: so that if you disable javascript you just get a nice tree
[01:54] creationix: only problem with a single file is that google can't index a particular page
[01:54] creationix: just the entire thing
[01:54] _ry: oh yeah, true
[01:55] creationix: I can modify my views to work on server or client and then serve both
[01:56] creationix: the url /api would be server-generated for google bots and /#api would be for client-side and offline rendering
[01:56] creationix: and the static version would have a javascript redirect to the dynamic one, so people coming in off the google search results would get sent to the faster versino
[01:59] creationix: (I wonder if I'm entering the "trying ti be smart" territory)
[02:00] isaacs: why not build it out as separate pages, old-school website-style, and then have a way to fetch just the content bit by doing ?ajax=1 or something, and do the show-hide stuff on the client if it supports it?
[02:00] _ry: creationix: i was just thinking that about my one html page :)
[02:00] creationix: seriously though, who's going to be reading javascript docs and not have a javascript enabled browser
[02:00] isaacs: then the single-page versions all link to the fancy pants site, but onlyonce you click a link, not a javascript redirect
[02:00] _ry: creationix: yeah, that's not a big concern
[02:01] rektide: chakrit1: sprocket is a tool written in ruby for managing javascript files.  it does dependency management, documentation into jsdoc, quite a few things.
[02:01] isaacs: the concern isn't non-js browsers.  it's that js redirects are SLOW
[02:01] rektide: :img-timeline:
[02:01] creationix: isaacs: I'm trying to avoid the requirement of pre-generating the html for most users
[02:01] _ry: creationix: have you thought about how to do the event tables?
[02:01] _ry: (they don't have to be tables)
[02:01] creationix: I've got them as ul's right now, but it's still hard to follow
[02:01] _ry: could just be a list 
[02:02] _ry: creationix: once you get markdown going well, you should patch me upstream - we can use "ronn" to generate the man page
[02:02] creationix: sweet, I was wondering if there was a markdown to man converter
[02:02] _ry: rtomayko wrote it
[02:02] _ry: http://rtomayko.github.com/ronn/ronn.1.html
[02:03] _ry: ^-- this btw, looks great
[02:03] _ry: what if we did something man page styled?
[02:03] creationix: I noticed the jQuery site is like man pages, just very different typography
[02:04] _ry: did you see http://www.jqapi.com/
[02:05] _ry: also looks nice
[02:06] _ry: creationix: (sorry to throw so many ideas at you at once) what if we did a single page thing, like what we have now, except sections can be expanded
[02:06] BryanWB has joined the channel
[02:06] _ry: i don't know. i'm really glad you're thinking about this though :)
[02:07] inimino: I *really* like having everything on one page
[02:07] BryanWB: creationix, the api docs look sweet
[02:07] _ry: inimino: me too because cmd+f works
[02:07] creationix: my current implementation is one ajax download per volume (api and changelog)
[02:07] inimino: the fact that the documentation fit on a single page (and I could scroll through it all) was a major plus when I first looked at node
[02:08] rektide: i need some kind of super-intelligent IRC client to tell me where this conversation started.
[02:08] rektide: :V
[02:08] _ry: there was a blog post about that today actually
[02:08] _ry: http://ozmm.org/posts/man_what.html
[02:08] RayMorgan has joined the channel
[02:08] _ry: "Well, man pages do the same thing. For instance, I've installed node.js via homebrew which means I can type man node at any time to see the docs to the entire node.js standard library. Seriously. Thanks mxcl and ryah."
[02:09] _ry: well not exactly the same but similar thought. it's nice to have stuff on one page.
[02:09] inimino: yeah
[02:09] rektide: at least five iphone apps dedicated to man pages, all of which cost money.  mwah ha ha. oh ho ho. wow.
[02:10] _ry: the API isn't going to get much bigger than this either - so maybe just having the single page be more browsable would be better?
[02:10] creationix: ahh, I got it, how about showing the full docs a single page on the right and leave the lhs navigation as it
[02:10] creationix: *as is
[02:10] creationix: the TOC on the left would just scroll to the right spot
[02:11] kriskowal: _ry i have found that node's single page man page is *very* nice, and am hoping to move in that direction for narwhal
[02:11] _ry: sounds neat
[02:11] JohnResig has joined the channel
[02:11] _ry: creationix: the really big question though, is how to deal with multiple versions of docs
[02:11] kriskowal: narwhal has a much larger module base though, so i'll probably have a glossary and a page per module when all is said and done
[02:11] _ry: just have a whole other site?
[02:12] creationix: I plan to put in a query parameter in the url where you can specify the version
[02:12] creationix: also it's decoupled enough that you would put in the url for a seperate project and pull data from it
[02:13] inimino: creationix: yeah, that sounds great (one page plus #fragids)
[02:13] creationix: so I could pull in the readme.markdown for Do and read the docs in the doc browser
[02:14] creationix: do hashes scroll inner divs or just the outer page?
[02:14] creationix: on static pages
[02:14] inimino: both
[02:14] inimino: if you have a fragment id inside something with a scroll bar in it, it will scroll both the outer page and the inner box
[02:15] inimino: creationix: so you could have something like a bar across the top with 'current version' 'older versions' ...?
[02:15] pedrobelo has joined the channel
[02:15] creationix: inimino: I was planning on putting a dropdown with a list of the versions
[02:16] inimino: sounds nice
[02:17] creationix: so how do I read in arbitrary man files without installing them to the system
[02:17] creationix: (let me guess, I should man man)
[02:17] inimino: man /path/to/file
[02:18] creationix: ok, easy enough
[02:19] derferman: creationix: after looking over the docs, I have a question. Why did you choose to implement two different scrolling boxes instead of just scrolling the whole page?
[02:19] creationix: deferman: I didn't design the site, I stole it from the jQuery giys
[02:19] creationix: I'm not sold on it yet
[02:20] derferman: yeah, either am I. The content is great, I am just unsure of the two scrolling columns
[02:20] wiber has joined the channel
[02:21] isaacs: creationix: where's your current build at?
[02:21] creationix: http://static.creationix.com/node_docs/browser/#api
[02:22] creationix: http://github.com/ry/node_docs
[02:22] isaacs: nice
[02:22] isaacs: but, yeah, two-scrolling-box is lame.
[02:22] isaacs: better to have just two columns
[02:22] pedrobelo has joined the channel
[02:23] derferman: agreed, I think it will make it much easier to use. I just hate having to scroll two different boxes
[02:23] deanlandolt has joined the channel
[02:23] creationix: why two columns that scroll together, then you'd not see the toc for most of the content
[02:23] creationix: you'd have to scroll to the top every time you wanted to change sections
[02:24] creationix: it's two boxes at http://nodejs.org/api.html
[02:24] creationix: only the left one doesn't ever have a scroll bar
[02:24] creationix: sucks if you're on a small screen
[02:24] _ry: yeah
[02:24] derferman: correct, and the left side takes up much less room
[02:25] derferman: I just feel like the right side box is constraining the content
[02:25] creationix: too much padding perhaps
[02:25] derferman: One of the things I like about the current node docs site is that the main docs take up the whole page
[02:25] derferman: no header, menu, or footer
[02:25] joshholt has joined the channel
[02:26] vviber has joined the channel
[02:26] creationix: true
[02:26] derferman: I do agree that lacking a scroll bar on the left is bad for smaller screens
[02:27] isaacs: derferman: but if it wasn't position:fixed, that'd be fine.
[02:27] derferman: isaacs: correct, that would fix the problem
[02:28] derferman: I guess it depends on how long the sections will be. If they are incredibly long, navigating back to the page might be tiresome
[02:29] bronson has joined the channel
[02:29] creationix: well, thanks for the input everyone
[02:30] wiber has joined the channel
[02:36] brainss has joined the channel
[02:38] Booster_ has joined the channel
[02:39] CIA-77: node: 03Ryan Dahl 07net2 * rc1a0ade 10/ (5 files in 4 dirs): Further net2 compatibilities - http://bit.ly/c8GtdP
[02:39] CIA-77: node: 03Ryan Dahl 07net2 * r0918bb2 10/ (4 files in 2 dirs): Adjust a few more tests to work with net2 - http://bit.ly/9ViELf
[02:47] pedrobelo has joined the channel
[02:57] pedrobelo has joined the channel
[03:03] indiefan has joined the channel
[03:04] wiber has left the channel
[03:06] kriszyp has joined the channel
[03:15] joshholt has joined the channel
[03:15] softdrink has joined the channel
[03:21] unom1 has joined the channel
[03:26] rictic has joined the channel
[03:28] bpot has joined the channel
[03:28] kriszyp has joined the channel
[03:28] RayMorgan has joined the channel
[03:28] mjijackson has joined the channel
[03:28] mjr_ has joined the channel
[03:28] rednul has joined the channel
[03:28] tlrobinson has joined the channel
[03:28] ivan has joined the channel
[03:34] halorgium has joined the channel
[03:36] derferman has joined the channel
[03:38] jed has joined the channel
[03:39] mikeal has joined the channel
[03:40] joshbuddy has joined the channel
[03:40] joshbuddy has joined the channel
[03:44] whoahbot has joined the channel
[03:47] dnolen_ has joined the channel
[03:53] bronson has joined the channel
[04:05] whoahbot has joined the channel
[04:13] mikeal has joined the channel
[04:23] jashkenas has joined the channel
[04:26] tilgovi has joined the channel
[04:39] PyroPete1 has joined the channel
[04:39] mikeal: i thought everyone finally stopped caring about coffeescript this week?
[04:39] mikeal: :P
[04:40] mjijackson_ has joined the channel
[04:41] jashkenas: Hey, it's a surprise to me just as much as it is to you.
[04:42] mikeal: i find it kind of entertaining how in to it everyone is, because this is not the first nor the last one of these
[04:42] mikeal: i just don't care about syntax
[04:43] mikeal: and i'm not willing to trade ease of debugging for syntax sugar
[04:45] mikeal: i'm surprised aza's pyscript doesn't have more buzz http://www.azarask.in/blog/post/making-javascript-syntax-not-suck/
[04:45] mikeal: it has support in Firefox via an extension already
[04:46] mikeal: i just don't want to see bloat in node
[04:46] mikeal: because that's the last thing it is right now
[04:47] mikeal: i have faith in ry tho :)
[04:47] jashkenas: amen to that. It's wonderful to run "cake test" and see everything compile and pass in 0.19 seconds.
[04:47] jashkenas: The equivalent hundreds of tests in Ruby would take lord knows how long.
[04:47] mikeal: i really can't believe how fast all of this is
[04:48] joshbuddy has joined the channel
[04:48] mikeal: i learned node and wrote a proxy in a couple hours
[04:48] mikeal: and it's faster than the proxy i worked on for 3 years in Python
[04:48] mikeal: it's so damn fast
[04:50] jashkenas: Well, it's funny to hear about "how into it everyone is", because I certainly don't hear anything out in the boonies here. Just bug tickets. Glad that the bay is picking it up.
[04:50] mikeal: it's not that a million people are using it
[04:51] mikeal: it's that all the right people are using it
[04:51] jashkenas: ah, high-society stuff, debutantes, polo players...
[04:51] mikeal: it's not ready for most people to use it yet
[04:51] mikeal: the average Python/Ruby developer can't deal with this rate of change in their platform
[04:53] mikeal: but it has this feeling of inevitability :)
[04:53] mikeal: everyone who is using it loves it
[04:53] mikeal: and the people working on it are concerned with real world problems instead of silly language semantics that don't really matter
[04:55] _ry: inevitability hm. 
[04:55] _ry: yeah, i think it nees a lot of work.
[04:56] _ry: *needs. inevitability is way too strong.
[04:57] mikeal: haha
[04:57] mikeal: it needs work while it's already one of the fastest things you can use :)
[04:58] jashkenas: The premise isn't too strong. Ubiquitous JS is going to happen. It's just a matter of who gets there first.
[04:58] mikeal: what's already built is ahead of Python and Ruby by a mile
[04:58] BryanWB: it is one thing to be a great tool, but every great tool needs a killer app to really take off
[04:58] BryanWB: as rails was for ruby
[04:58] mikeal: that'll come eventually
[04:58] _ry: jashkenas: i agree 
[04:58] mikeal: i don't have much faith in the current web frameworks on the list
[04:59] mikeal: but something awesome will come around
[04:59] _ry: but i think it's a year out before we start seeing real websites on js
[04:59] mikeal: there are already some huge apps on CouchDB which are basically in SSJS
[05:00] mikeal: there is a Meebo app with a million installs
[05:00] mikeal: and the BBC is using it
[05:00] mikeal: does some insane number of requests per day
[05:01] mikeal: but the only reason that works is because the view server isn't allowed to do IO
[05:01] RayMorgan has joined the channel
[05:01] mikeal: it can only do processing
[05:01] mikeal: node can do everything
[05:01] _ry: (except parse xml ;))
[05:01] mikeal: someone just needs to make it dead simple and there are far less limits
[05:01] mikeal: spidermonkey has e4x :)
[05:02] mikeal: which i hate
[05:02] mikeal: but it's something
[05:02] mikeal: isaacs sax parser looks good
[05:02] _ry: it needs work
[05:02] RayMorgan has joined the channel
[05:02] technoweenie has joined the channel
[05:02] brainss has joined the channel
[05:02] _ry: but yeah, could be really good
[05:02] mikeal: everything needs work
[05:03] mikeal: right now i just see a lot of promise
[05:03] mikeal: even tho we dropped Promises
[05:03] _Pilate has left the channel
[05:03] mikeal: bu dum chinng
[05:04] Pilate has joined the channel
[05:05] _ry: 
ACTION is writing sexy idle timeout hotness

[05:06] _ry: it's so sexy and hot
[05:06] mikeal: sounds hot
[05:06] _ry: writing stuff in js is so much more pleasurable than c
[05:06] jashkenas: idle timeouts?
[05:08] _ry: for sockets
[05:13] binary42 has joined the channel
[05:14] morgan: someone sent a pure JS html5 parser to the env-js mailing list today
[05:14] morgan: http://groups.google.com/group/envjs/browse_thread/thread/61274fed3458c120?hl=en
[05:14] morgan: only briefly looked at it, but with only a couple changes it would load in node
[05:15] morgan: didn't get as far as parsing anything
[05:23] _ry: morgan: oh cool
[05:28] kennethkalmer has joined the channel
[05:28] jashkenas has left the channel
[05:30] nefD has joined the channel
[05:32] mattly has joined the channel
[05:43] jan____ has joined the channel
[05:48] jed has joined the channel
[05:49] aryounce has joined the channel
[05:53] isaacs has joined the channel
[05:53] jan____ has joined the channel
[05:57] neytema has joined the channel
[05:57] jan____ has joined the channel
[05:59] morgan has joined the channel
[05:59] jan____ has joined the channel
[06:01] jan____ has joined the channel
[06:02] brapse has joined the channel
[06:02] jan____ has joined the channel
[06:03] jan____ has joined the channel
[06:11] shansen has joined the channel
[06:24] tlrobinson_ has joined the channel
[06:25] JimBastard has joined the channel
[06:26] steadicat has joined the channel
[06:27] emyller has joined the channel
[06:36] dnolen has joined the channel
[06:41] JimBastard: anyone an nginx ninja? trying to do NginxHttpRewriteModule with a proxy_pass to node
[07:04] mikeal: i haven't used nginx for a while
[07:04] mikeal: and to be honest
[07:04] mikeal: i never really understood how the rules worked
[07:04] mikeal: why not just run node on port 80?
[07:04] mikeal: it's pretty good at concurrency :)
[07:05] Pilate: because its probably being used by other domains on the box :P
[07:05] JimBastard: unsuccessful troll is not troll
[07:05] JimBastard: we have a bunch of static content
[07:05] JimBastard: as much as i love flooding node with requests
[07:05] JimBastard: we are using nginx for the static stuff
[07:05] mikeal: node *should* be better at that
[07:06] JimBastard: impossible
[07:06] mikeal: nginx uses blocking file IO
[07:06] mikeal: :)
[07:06] mikeal: nonblocking socket IO, but file IO is blocking
[07:06] JimBastard: really just have no idea how to configure nginx
[07:06] JimBastard: and my russian is terrible
[07:06] mikeal: haha
[07:06] Pilate: are you using it already?
[07:06] JimBastard: so the irc isnt being friendly
[07:06] mikeal: i got it working once
[07:07] JimBastard: its up and running right now
[07:07] mikeal: and even tho it was painful, the rules still seemed simpler than apache
[07:07] JimBastard: and the proxy_pass is passing to node
[07:07] JimBastard: working well
[07:07] Pilate: so what are you trying to add?
[07:07] JimBastard: i need to do a rewrite to my proxy_pass server for 404s, passing the path
[07:07] JimBastard: so if mydomain.com/blahfoo comes up
[07:07] JimBastard: it proxy_pass to 127.0.0.1:8080/blahfoo
[07:08] JimBastard: but only if its a 404, has to check if its a static file path first
[07:08] mikeal: you can refer to their amazing documentation :P
[07:08] mikeal: i don't know if you can do that
[07:08] mikeal: have something be a file path but if it's a 404 be something else
[07:08] jed has joined the channel
[07:13] Pilate: http://wiki.nginx.org/NginxHttpCoreModule#error_page
[07:13] Pilate: it mentions 404 fallback
[07:24] derferman has joined the channel
[07:25] brapse has joined the channel
[07:31] BryanWB has joined the channel
[07:35] javajunky has joined the channel
[07:36] kjeldahl has joined the channel
[07:36] javajunky1 has joined the channel
[07:37] javajunky1 has left the channel
[07:45] kennethkalmer has joined the channel
[07:45] kennethkalmer has joined the channel
[07:54] unom1 has joined the channel
[08:04] markwubben has joined the channel
[08:04] piranha has joined the channel
[08:05] _ry: JimBastard: i posted an nginx config to the list once
[08:07] JimBastard: hi _ry , i didnt see what you posted but im assuming its just the basic proxy_pass?
[08:07] JimBastard: i do have node and nginx playing nice, im just trying to get fancy
[08:11] jed has joined the channel
[08:15] rudebwoy has joined the channel
[08:26] rudebwoy has joined the channel
[08:30] _ry: JimBastard: oh okya
[08:30] _ry: yeah it was just a simple thing
[08:31] BryanWB has joined the channel
[08:32] JimBastard: thank you though, ill figure it out im sure
[08:35] paulca has joined the channel
[08:41] rtomayko has joined the channel
[08:43] christkv has joined the channel
[08:44] spoob has joined the channel
[08:45] Mandar has joined the channel
[08:59] mAritz has joined the channel
[09:33] bronson has joined the channel
[09:56] felixge has joined the channel
[09:56] felixge has joined the channel
[10:02] CIA-77: node: 03Ryan Dahl 07net2 * r252673e 10/ (lib/net.js src/node_timer.cc test/pummel/test-tcp-timeout.js): net2 timeouts work - http://bit.ly/c5Ie1s
[10:04] _ry: fuck - forcing update to net2
[10:04] Tim_Smart has joined the channel
[10:07] christkv: _ry: any direction on what api document generator you are planning to use ?
[10:14] mcarter has joined the channel
[10:27] ithinkihaveacat has joined the channel
[10:35] christkv: is it possible to use http via a proxy from node ?
[10:38] hassox has joined the channel
[10:56] MattJ has joined the channel
[11:07] bronson has joined the channel
[11:09] bru has joined the channel
[11:22] Tim_Smart has joined the channel
[11:30] rtomayko has joined the channel
[11:31] tlrobinson has joined the channel
[11:31] derferman has joined the channel
[11:41] piranha has joined the channel
[12:09] pdelgallego has joined the channel
[12:20] paulca has joined the channel
[12:37] alex-desktop has joined the channel
[12:48] bru_ has joined the channel
[12:51] dnolen has joined the channel
[12:52] sveimac_ has joined the channel
[12:55] sveimac_ has joined the channel
[13:09] mAritz has joined the channel
[13:13] pjb3 has joined the channel
[13:13] bru_ has left the channel
[13:17] dnolen has joined the channel
[13:28] paulca has joined the channel
[13:29] kjeldahl has joined the channel
[13:44] siculars_ has joined the channel
[13:49] pmuellr has joined the channel
[13:50] kriszyp has joined the channel
[13:51] dnolen_ has joined the channel
[13:52] kjeldahl has joined the channel
[13:55] michaelk^ has joined the channel
[13:57] BryanWB has joined the channel
[13:59] ithinkihaveacat has joined the channel
[14:01] jspiros has joined the channel
[14:03] kjeldahl has joined the channel
[14:09] jherdman has joined the channel
[14:14] jed_ has joined the channel
[14:17] gf3 has joined the channel
[14:20] chakrit has joined the channel
[14:23] chakrit1 has joined the channel
[14:28] Booster has joined the channel
[14:32] sveimac has joined the channel
[14:36] n8o__ has joined the channel
[14:38] bronson has joined the channel
[14:42] binary42 has joined the channel
[14:42] binary42 has joined the channel
[14:43] stepheneb has joined the channel
[14:44] aryounce has joined the channel
[14:52] Sidnicious has joined the channel
[14:52] joshbuddy has joined the channel
[14:52] joshbuddy has joined the channel
[14:56] pdelgallego has joined the channel
[15:15] BryanWB: anyone know if v8 continue to make significant speed increases?
[15:15] BryanWB: or if speed increase will start marginally diminishing?
[15:16] ashb: from what i gather JagerMonkey is catching up
[15:16] alex-desktop has joined the channel
[15:16] ashb: tracing wins in for certain usage patterns and by a long way, and where it doesn't then the general JIT approach will get you a very large way to v8's performance
[15:17] ashb: (you still loose a bit initially when tracing is attempted)
[15:17] BryanWB: ashb, that's fine if jagermonkey catches up
[15:17] BryanWB: ashb, i guess my question is whether js interpreters will continue to make big strides
[15:17] BryanWB: in general
[15:17] ashb: as far as i'm concernec this kind of competition between v8, JS Core and Spidermonkey can only be a good thing for all of us
[15:18] ashb: i think not.
[15:18] BryanWB: ashb, definitely, just curious if we will continue to see big speed increases or hit a hard limit
[15:18] ashb: now that they are all JITting it is more likely to just get incremental increases
[15:19] ashb: which isn't to rule out other big jumps, but they aren't as likely to as big as interpreted -> jitted
[15:19] BryanWB: ashb, sure
[15:19] softdrink has joined the channel
[15:19] aryounce: BryanWB: Depends I guess. I was at I/O last year and the V8 team said they think they still have a long way to go
[15:20] BryanWB: ashb, have any idea when we will be able to run pre-jited code rather than having to jit every time?
[15:20] ashb: BryanWB: none. I dont follow v8 dev at all
[15:20] BryanWB: like .pyc files w/ python
[15:20] BryanWB: aryounce, good to hear
[15:20] aryounce: BryanWB: at the very least they have a laundry list of improvements over their initial attempt that I'm sure have been slowly making it in
[15:21] Mandar has left the channel
[15:21] BryanWB: type annotations are in ecma 5 correct?
[15:21] BryanWB: code using annotations should be lots faster
[15:21] ashb: BryanWB: no - types was ES4
[15:22] BryanWB: ashb, ah so no type annotations
[15:22] ashb: no.
[15:22] ashb: not until ES next+1 anyway
[15:31] kriszyp_ has joined the channel
[15:31] ashb: aryounce: did they mention what kind of things they were?
[15:31] aryounce: ashb: It's been so long, let me see if I can find the youtube video
[15:31] ashb: nm
[15:31] ashb: not curious enough to watch a video :)
[15:32] aryounce: ashb: ha, there was a Q&A at the end too
[15:32] aryounce: ashb: which may not have made it into the video
[15:32] ashb: :)
[15:32] aryounce: but there was a ton of good info on how V8 worked
[15:32] aryounce: the presentation was from one of the first engineers to work on it
[15:35] qFox has joined the channel
[15:57] sveimac has joined the channel
[15:57] cloudhead has joined the channel
[16:03] stepheneb has joined the channel
[16:03] dnolen has joined the channel
[16:04] technoweenie has joined the channel
[16:10] alexiskander has joined the channel
[16:11] shansen has joined the channel
[16:19] r11t has joined the channel
[16:19] bpot has joined the channel
[16:19] codeswing has joined the channel
[16:20] christkv: _ry: the v8 issue http://code.google.com/p/v8/issues/detail?id=636 has been fixed in v2.1.3
[16:25] steadicat has joined the channel
[16:25] brapse has joined the channel
[16:29] _ry: christkv: nice
[16:29] _ry: christkv: api document generator?
[16:30] jed_ has joined the channel
[16:30] christkv: yeah the replacement for ascidoc
[16:30] _ry: oh, i think we're going to do it by hand
[16:30] _ry: creationix is working on it
[16:30] creationix has joined the channel
[16:30] _ry: markdown, i think he's using
[16:31] _ry: oh there he is :)
[16:31] _ry: christkv: http://github.com/ry/node_docs
[16:31] creationix: hey
[16:31] ashb: markdown i hear you say?
[16:32] ashb: booo showdown
[16:32] ashb: creationix: you should totally be using http://github.com/evilstreak/markdown-js/
[16:33] ashb: (i'll tidy it up so its a single file tonight like it can/shoudl be)
[16:34] creationix: ashb: how does it compare to showdown
[16:34] ashb: creationix: its not a port of the buggy as well markdown.pl ;)
[16:34] christkv: something like rdoc would be nice
[16:34] ashb: its written to be easily extensible to add dialects
[16:35] ashb: and to give you an IR that you can tweak with/inspect if oyu want
[16:35] creationix: ashb: that would be great
[16:35] ashb: oh. and produces well-formed HTML
[16:35] paulca has joined the channel
[16:35] whoahbot has joined the channel
[16:36] ashb: creationix: see the readme for our reasons for writing it
[16:36] ashb: but basically: 1) not tied to DOM, 2) valid, 3) easy to extend
[16:36] ashb: 4) WITH TESTS
[16:38] creationix: sounds good, I may need to extend the dialact a but
[16:38] creationix: *bit
[16:38] christkv: I'm just hoping some sort of agreement comes out so I can auto-generate the api docs
[16:38] christkv: what do you plan to use for node itself ?
[16:38] davidjrice has joined the channel
[16:38] ashb: creationix: http://github.com/evilstreak/markdown-js/blob/master/lib/markdown.js#L993
[16:39] ashb: creationix: through to L1108
[16:41] creationix: nice
[16:42] ashb: (technically the document meta is part of PHP Extra i think, but eh)
[16:43] voodootikigod_ has joined the channel
[16:48] RayMorgan has joined the channel
[16:50] ashb: creationix: off out for a while - le me know if oyu have any problems
[16:54] creationix: alright, thanks
[16:55] bpot has joined the channel
[16:58] Booster has joined the channel
[17:14] michaelk^ has joined the channel
[17:15] cedricv has joined the channel
[17:16] christkv: I'm putting together a small example app using express + the mongo driver that will run on nodeblogs.com at some point. here's the first screenshot http://skitch.com/christkv/n3epi/node-blogs
[17:24] creationix: christkv: so how will it work? pull in rss feeds and display them to the user
[17:25] creationix: it does look cool though
[17:26] aguynamedben has joined the channel
[17:26] kriskowal has joined the channel
[17:27] creationix: hmm, I'd better leave before the storm gets up there.  It's been hailing and everything here
[17:28] bronson has joined the channel
[17:29] kennethkalmer has joined the channel
[17:30] brandon_beacher has joined the channel
[17:31] christkv: yeah it pulls in rss, github and twitters with #nodejs
[17:31] creationix: whoops, my last message was for #thechangelog, wrong channel
[17:36] aguynamedben has joined the channel
[17:43] dandean has joined the channel
[17:50] christkv has joined the channel
[17:50] bpot has joined the channel
[17:55] bpot has joined the channel
[17:59] davidjrice has left the channel
[18:02] mjijackson_ has left the channel
[18:02] mjijackson has joined the channel
[18:05] gf3 has joined the channel
[18:06] gf3 has joined the channel
[18:10] stepheneb has joined the channel
[18:11] neytema has joined the channel
[18:29] mikeal has joined the channel
[18:30] stepheneb has joined the channel
[18:36] aho has joined the channel
[18:38] RayMorgan_ has joined the channel
[18:49] derferman has joined the channel
[18:49] joshholt has joined the channel
[18:49] atmos: what's the opposite of JSON.stringify ? 
[18:49] technoweenie: JSON.parse isnt it
[18:50] atmos: thanks
[18:50] atmos: didn't expect a rubyish api to work for some reason :)
[18:50] technoweenie: iirc its the original json lib
[18:51] inimino: indeed
[18:51] technoweenie: so ruby copied it
[18:51] inimino: `g json2 js
[18:51] gbot2: JSON in JavaScript - http://www.json.org/js.html
[18:51] technoweenie: ruby's json gem, rather
[18:52] inimino: Doug Crockford wrote the original implementation, and now it's standardized in ES5 with the same API
[18:53] _ry: v8 is insane: 
[18:53] _ry:  195 files changed, 10342 insertions(+), 3251 deletions(-)
[18:53] inimino: heh
[18:53] mikeal has joined the channel
[18:53] _ry: they add like 8000 lines of c++ every release
[18:53] inimino: on that note, I think I'll go write some code
[18:56] CIA-77: node: 03Ryan Dahl 07master * r073947c 10/ (195 files in 13 dirs): Upgrade V8 to 2.1.3 - http://bit.ly/bmzFqK
[18:57] cedricv has joined the channel
[18:59] mikeal: is most of the code they are writing still to support ARM?
[18:59] JimBastard has joined the channel
[19:00] _ry: mikeal: i don't think so
[19:01] _ry: they fixed the fucking gcc 4.4 bug
[19:01] _ry: finally!
[19:01] _ry: (by disabling -fstrict-aliasing
[19:01] derferman has joined the channel
[19:02] _ry: "bug"
[19:03] _ry: oh man. wtf. http://buildbot.nodejs.org:8010/builders/Linux%202.6.32/builds/68/steps/shell_1/logs/stdio
[19:03] _ry: :(
[19:03] felixge: I wish v8 changelogs were better
[19:03] felixge: _ry: that means they did not fix it?
[19:03] CIA-77: node: 03Ryan Dahl 07master * r70b2a44 10/ deps/v8/SConstruct : Remove -Werror from V8 -
[19:03] CIA-77: node: 03Ryan Dahl 07net2 * r70b2a44 10/ deps/v8/SConstruct : Remove -Werror from V8 - http://bit.ly/abPX2Q
[19:04] CIA-77: node: 03Ryan Dahl 07net2 * r777e1a6 10/ deps/v8/SConstruct : Merge branch 'master' into net2 - http://bit.ly/9NMfdL
[19:04] _ry: nope
[19:04] _ry: or maybe that's a different issue
[19:05] felixge: _ry: are you using the default build bot theme or something custom?
[19:06] ollie has joined the channel
[19:07] javajunky has joined the channel
[19:10] joshbuddy has joined the channel
[19:11] creationix has joined the channel
[19:11] CIA-77: node: 03Ryan Dahl 07net2 * rdd21a4f 10/ src/node.cc : Remove the 'Error: (no message)' exceptions print stack trace instead - http://bit.ly/9AGbcX
[19:11] CIA-77: node: 03Standa Opichal 07net2 * rc2c0cfb 10/ bin/node-waf : Making sure node-waf finds its real bindir even when executed through a symlinked path. - http://bit.ly/drPeED
[19:11] CIA-77: node: 03Ryan Dahl 07net2 * r073947c 10/ (195 files in 13 dirs): Upgrade V8 to 2.1.3 - http://bit.ly/bmzFqK
[19:11] CIA-77: node: 03Ryan Dahl 07net2 * r7eb1264 10/ (196 files in 14 dirs): Merge branch 'master' into net2 - http://bit.ly/dzTqLy
[19:13] _ry: felixge: i think the default
[19:13] _ry: felixge: well the homepage is using the node colors
[19:14] felixge: _ry: cool
[19:14] felixge: their homepage still shows old ugly screenshots
[19:14] felixge: I think the new theme must come from google chrome's buildbut
[19:14] felixge: * buildbot
[19:14] _ry: yeah buildbot in gereral could use a graphic artist
[19:15] _ry: looks like shit
[19:15] JimBastard: github is fucked up again i think
[19:15] joshbuddy_ has joined the channel
[19:15] JimBastard: anyone else getting pack errors / device full errors
[19:15] michaelk^ has joined the channel
[19:16] softdrink has joined the channel
[19:19] christkv_ has joined the channel
[19:22] tilgovi has joined the channel
[19:24] javajunky: JimBastard: I'm pulling from it ok, not pushed though if that helps at all
[19:24] JimBastard: yeah push is broken again
[19:24] JimBastard: same issue a few weeks ago where device is reporting full
[19:27] gwoo has joined the channel
[19:29] maushu has joined the channel
[19:34] mikeal has joined the channel
[19:35] dnolen has joined the channel
[19:35] voodootikigod: http://jsconf.posterous.com/the-community-that-cares
[19:44] kjeldahl_ has joined the channel
[19:51] javajunky has joined the channel
[19:52] brandon_beacher has joined the channel
[19:53] aho: when is the next jsconf in eu?
[19:55] JimBastard: not sure
[19:55] JimBastard: does anyone know any sponsers with extra tickets? or anyone? i couldnt afford to buy them when first when on sale, now sold out
[19:55] JimBastard: i really wanted to do a lightning talk for this node project im working on with a few people
[19:56] JimBastard: also get drunk with ry
[19:56] mikeal: it's in september
[19:56] mikeal: in Berlin
[19:56] mikeal: tickets for JSConf EU haven't gone out yet
[19:56] JimBastard: i mean the US one
[19:56] JimBastard: in DC
[19:56] JimBastard: in april
[19:56] mikeal: yeah, good luck
[19:57] JimBastard: lol
[19:57] mikeal: i've been trying to get a ticket for weeks
[19:57] JimBastard: i would have bought one the day they went on sale but im not exactly big pimping like that
[19:57] mikeal: i was on vacation when they all went
[19:57] JimBastard: it took me a while to get the money, as i was unemployeed
[19:57] JimBastard: yeah
[19:57] mikeal: that's what credit cards are for :)
[19:58] JimBastard: ohh im waaay past that
[19:58] JimBastard: ever since i got sued for 550k my credit has been not so great
[19:59] mikeal: how did you get sued for a half a million dollars?
[19:59] JimBastard: a laywer and a plantiff
[19:59] JimBastard: plaintiff
[19:59] indiefan: that'll do it
[19:59] JimBastard: i released this viral video in 2001 or so
[19:59] JimBastard: it got a little out of hand
[20:00] indiefan: was it bum fights?
[20:00] mikeal: the virus started killing people
[20:00] JimBastard: no comment
[20:00] mikeal: was it the swine flu or youtube
[20:00] mikeal: s/or/of
[20:00] mikeal: that would have been funnier if i could type
[20:00] JimBastard: this was 2001, so kazaa and word of mouth
[20:01] JimBastard: im actually not allowed to talk about it, gag order
[20:01] mikeal: how do you get sued for real money tho, there is no money on the internet :)
[20:01] felixge: lol
[20:01] JimBastard: if you google JimBastard enough you might find something
[20:01] indiefan: 
ACTION checks if once is enough

[20:02] JimBastard: there we go
[20:02] JimBastard: http://nymag.com/news/features/27341/index5.html
[20:02] mikeal: HAHA!
[20:04] JimBastard: yeah im like a C list internet celebrity, i think that entitles me to a discount on movie tickets
[20:04] JimBastard: and free jsconf tickets
[20:04] JimBastard: ^_^
[20:07] piranha has joined the channel
[20:11] javajunky1 has joined the channel
[20:11] rictic has joined the channel
[20:12] drostie has joined the channel
[20:12] pdelgallego has joined the channel
[20:14] creationix has left the channel
[20:16] mattly has joined the channel
[20:19] voodootikigod has joined the channel
[20:21] piranha has joined the channel
[20:25] JAAulde has joined the channel
[20:25] creationix has joined the channel
[20:27] pmuellr has joined the channel
[20:28] orlandov: can anyone familiar with the HTML5 async Web DB explain to me how it's possible for openDatabse to be asynchronous, open a db and return the db handle in one call?
[20:28] halorgium: orlandov: the handle might not yet be connected, but a pointer to something which will become connected
[20:28] joshholt has joined the channel
[20:29] brainss has joined the channel
[20:29] orlandov: halorgium: hmmm i suppose
[20:30] orlandov: i guess you'd have to build a backlog of query calls and start executing them once it does connect
[20:30] halorgium: orlandov: in general, you'd disallow queries to be made until the handle was ready
[20:31] orlandov: ugh... my gut reaction is to not like that
[20:31] orlandov: i wish it was just callbacks like node
[20:32] halorgium: https://gist.github.com/f48c7b54461bb74d303f something like that
[20:32] halorgium: is it already impl'd?
[20:33] orlandov: halorgium: i'm trying to implement a sqlite3 driver for node with the HTML5 async db interface
[20:33] orlandov: i have a low level api and now i'm trying to make it semi-standard-like
[20:34] isaacs has joined the channel
[20:36] orlandov: halorgium: i was going to do that at first (emitting events) but ryah convinced me it was probably better to just use callbacks
[20:36] halorgium: orlandov: similar end result
[20:36] halorgium: callbacks are easier to handle
[20:37] orlandov: yeah, and there's more or less 1:1 relationship between call and result
[20:37] orlandov: (if you ignore sqlite3 stepping through rows)
[20:38] orlandov: i think i'm just going to make this loosely based on the spec
[20:38] orlandov: bbl- runch time
[20:39] joshholt has joined the channel
[20:43] tlrobinson_ has joined the channel
[20:44] Tim_Smart has joined the channel
[20:46] JimBastard: morning Tim_Smart 
[20:46] Tim_Smart: Ello JimBastard 
[20:46] JimBastard: My friends call me JimBastard, you may call me Mr. Bastard
[20:46] JimBastard: :p
[20:50] rektide: tell me you have royal ancestery, you royal bastard
[20:51] kkaefer has joined the channel
[20:51] gwoo has joined the channel
[20:55] brainss has joined the channel
[20:57] RayMorgan has joined the channel
[20:57] pedrobelo has joined the channel
[20:58] joshbuddy has joined the channel
[21:06] stepheneb has joined the channel
[21:06] sveisvei: JimBastard: Make my autocomplete autocomplete to Mr.Bastard if thats the rule ;-)
[21:09] derferman has joined the channel
[21:26] maushu: Where is the "require" code? Is it in javascript or in node's sauce?
[21:28] felixge: maushu: src/node.js
[21:28] maushu: Ah, thanks.
[21:35] whoahbot has joined the channel
[21:36] brandon_beacher has joined the channel
[21:38] mattly has joined the channel
[21:39] cpleppert has joined the channel
[21:40] maushu: Is there a way to prevent node from using node.js? Like start a process without it.
[21:42] mikeal: you could just start v8
[21:42] maushu: ...true.
[21:43] mikeal: what from node you want that isn't part of v8?
[21:43] maushu: Trying my own sandbox.
[21:44] maushu: But I think even v8 has some functions I wouldn't want.
[21:44] mikeal: you might want to look at v8 contexts
[21:44] maushu: Okay.
[21:45] mikeal: http://code.google.com/apis/v8/embed.html#contexts
[21:46] maushu: Ah man, I would like to avoid c++.
[21:48] isaacs: anyone know how i'm supposed to be writing the request body for a PUT request?
[21:48] indiefan has joined the channel
[21:48] isaacs: I'm getting an error that says the socket is not open for writing.
[21:48] isaacs: i swear this worked a while ago...
[21:50] maushu: isaacs, are you closing the socket outside the callback?
[21:50] maushu: This happened to me once.
[21:50] isaacs: i thought write() was sync?
[21:50] maritz has joined the channel
[21:50] mikeal: nope
[21:50] isaacs: client.request(...);  client.write(data, encoding);  client.close()
[21:50] maushu: isaacs, there is no sync here. nouh.
[21:50] isaacs: maushu: according to the docs, it's sync
[21:51] isaacs: it's a socket write, not fs
[21:51] maushu: The docs lie.
[21:51] mikeal: haha
[21:51] maushu: They lie!
[21:51] JimBastard: the docs are a lie
[21:51] mikeal: isaacs: you're proxying a request right
[21:52] isaacs: mikeal: no, making a PUT
[21:52] isaacs: puts(require("http").ClientRequest.prototype.write.toString())
[21:52] isaacs: that doesn't show a cb param
[21:52] isaacs: it's quite clearly synchronous
[21:52] mikeal: you're creating a put
[21:52] isaacs: mikeal: yeah
[21:52] mikeal: request.write()
[21:52] mikeal: request.end()
[21:53] mikeal: er
[21:53] mikeal: no
[21:53] mikeal: request.close()
[21:53] isaacs: request.close()
[21:53] isaacs: right
[21:53] mikeal: you said earlier you were calling client.write
[21:53] mikeal: but it's not on the client object, it's on the request object
[21:53] isaacs: mikeal: http://gist.github.com/328466
[21:53] mikeal: .close() will flush whatever is in the write buffer
[21:54] mikeal: nop[e
[21:54] mikeal: that's you're problem
[21:54] isaacs: i see
[21:54] mikeal: client.request() returns a request object
[21:54] mikeal: and you need to call .write() and .close() on that object
[21:54] mikeal: and you can keep around that client object to do a keep-alive connectino
[21:55] isaacs: i see
[21:56] isaacs: and the "response" event is emitted by the client? or the request?
[21:56] mikeal: the request
[21:56] mikeal: i'll just edit this
[21:56] isaacs: i got it.  kewl.  getting a 500 error now, so that's progress
[21:56] mikeal: http://gist.github.com/328472
[21:57] isaacs: what threw me is that .write() isn't chainable
[21:57] mikeal: chainable?
[21:57] mikeal: you can write chunks
[21:57] isaacs: doesn't return the ClientRequest object, so you can't do req.write(foo).close()
[21:57] mikeal: that's because it's suppose to be used for chunks
[21:57] isaacs: well, sure, but... still...
[21:57] mikeal: you can call it multiple times before .close()
[21:57] isaacs: that's just uggo
[21:58] isaacs: you'd still be able to do req.write().write().write().close()
[21:58] isaacs: if you didn't use the returned object, it's not so bad.
[21:58] isaacs: but makes the simple case simpler
[21:58] mikeal: that would work if this was jQuery :P
[21:58] mikeal: i think it's assumed that you'll actually have something like
[21:59] mikeal: read('file', function (chunk) { request.write(chunk) }, function (error) { request.close() } )
[22:00] isaacs: ok, gotta run.  i'll be back in a bit.  i'm creating npm publish url/to/tarball to both create (if it's the first time) and push a new version (if it's not)
[22:00] mikeal: BTW
[22:00] mikeal: github really needs to fix their tarballs on master
[22:00] isaacs: SRSLY!!
[22:00] mikeal: i had to run the npm installer like 6 times the other day before it worked
[22:00] isaacs: at some point, i'm thinking it'd be good for the registry to save the tarballs as attachment
[22:00] isaacs: s
[22:01] mikeal: couchdb is good at that
[22:01] isaacs: so you can publish a local tarball, and it'll upload it, then use the registry's attachment as the dist-url
[22:01] mikeal: tags are a lot more reliable on github tho
[22:01] mikeal: that will be good for people who don't use github
[22:01] mikeal: all 5 of them :P
[22:01] isaacs: mikeal: no, just one.  inimino
[22:08] derferman has joined the channel
[22:09] javajunky has joined the channel
[22:10] jashkenas has joined the channel
[22:11] felixge has joined the channel
[22:11] felixge has joined the channel
[22:12] inimino: haha
[22:15] JimBastard: FUCK YOU ROR
[22:16] JimBastard: RACK SESSIONS IS DEFINTELY NOT IN MY JOB DESCRIPTION
[22:17] isaacs_mobile has joined the channel
[22:20] felixge: JimBastard: lol, I had this thought twice today as well. I really need to say 'nay' more often :)
[22:20] isaacs_mobile: Happy international day of awesomeness
[22:20] felixge: isaacs: was beer invented today?
[22:20] maushu: Can anyone recommend a recent javascript book?
[22:20] Sidnicious: JavaScript: The Good Parts?
[22:21] mjijackson: maushu: all JavaScript books suck
[22:21] felixge: Sidnicious: +1
[22:21] softdrink: http://eloquentjavascript.net/
[22:21] Sidnicious: Secrets of the JavaScript Ninja
[22:21] felixge: mjijackson: the good parts does not suck : )
[22:21] isaacs_mobile: felixge: yes it was!
[22:21] JimBastard: yo mjijackson - too soon
[22:21] mjijackson: 
ACTION has never read the good parts

[22:21] JimBastard: i saw your website too , very distasteful you shouldnt be mocking micheal jackson like that
[22:21] JimBastard: ;-)
[22:21] isaacs_mobile: felixge: but it was just to celebrate the day, not the cause of the holiday
[22:21] maushu: Sidnicious, that was one was released?
[22:22] mjijackson: correction: every JavaScript book I've ever read sucks ;)
[22:22] maushu: *last
[22:22] JimBastard: shit i dont even know how to read
[22:22] felixge: mjijackson: a picture says it all: http://static.debuggable.com.s3.amazonaws.com/img/js-good-parts.jpg
[22:22] Sidnicious: maushu: Hey, you said recent. I'm just getting a jump on all the other recommendations ;)
[22:23] mjijackson: JimBastard: You should learn how to spell "Michael". It's only the most common name in the history of mankind. ;)
[22:23] cloudhead has joined the channel
[22:23] maushu: felixge, outch.
[22:23] Sidnicious: All right, I haven't read it, but it's probably going to be good.
[22:23] JimBastard: you'll have to excuse me if my flaming of you is limited to the span of a 3 second alt-tab, i just dont know you well enough yet
[22:23] felixge: mjijackson: the history of man kind where the chinese did not exist? :)
[22:24] orlandov: mjijackson: actually i think mohammed is the most common name on earth
[22:24] jashkenas has left the channel
[22:24] mjijackson: 
ACTION has never met a Chinaman named Michael. point taken

[22:24] Sidnicious has left the channel
[22:24] Sidnicious has joined the channel
[22:24] Sidnicious: damn these buttons
[22:25] mjijackson: 
ACTION needs to stop making blanket statements

[22:26] orlandov: mjijackson: dude chinaman is not the preferred nomenclature
[22:26] JimBastard: yeah they perfer "Jackie Chan"
[22:26] JimBastard: prefer
[22:27] mjijackson: orlandov: no offense intended, really
[22:27] orlandov: mjijackson: none taken, it's a big lebowski quote :)
[22:27] isaacs_mobile: This non-PC convo is way more awesome than the line I'm waiting in.
[22:27] orlandov: http://www.imdb.com/title/tt0118715/quotes?qt0464758
[22:27] mjijackson: Seems the Wikipedia page on Chinaman was written by someone who takes great offense at the term's usage
[22:28] isaacs_mobile: I don't grok getting offended at terms like that.
[22:28] cpleppert has joined the channel
[22:29] isaacs_mobile: If you called me an Americaman I wouldn't care
[22:29] maushu: mjijackson, NO NPV? HORRIBLE.
[22:30] mjijackson: maushu: net present value? huh?
[22:30] maushu: mjijackson, neutral point of view, you silly person.
[22:32] mjijackson: maushu: i don't think i've ever read a truly NPV, anywhere
[22:32] maushu: Of course.
[22:32] mjijackson: Especially not on wikipedia
[22:33] maushu: Like my gramps liked to quote, "History is written by the victors.".
[22:35] mjijackson: maushu: true. about the js book though, what are you hoping to get out of it?
[22:35] maushu: contexts
[22:37] mjijackson: JavaScript execution contexts? Like variable scope and that?
[22:37] maushu: Yes.
[22:38] ashb: read the spec
[22:38] maushu: Life is short.
[22:38] gsf: maushu: yeah, the good parts boils that down pretty well
[22:39] mjijackson: https://developer.mozilla.org/en/Core_JavaScript_1.5_Reference/Statements/var
[22:39] mjijackson: i've always wondered why Mozilla serves their js docs over a secure connection
[22:40] maushu: mjijackson, they are secret.
[22:40] ashb: cos they use some abomination of comercial mediawiki instance i thinnk
[22:41] maushu: Does v8 support const?
[22:41] ashb: no. its not ES5 either
[22:41] ashb: its an ES-harmony proposal tho
[22:49] mjijackson: Arrrrgh! I need to stop playing with Node.js! I put off everything else... it's addicting.
[22:49] michaelk^ has joined the channel
[22:54] paulca has joined the channel
[22:59] rektide: need a linux vm for this desktop, so i can run node on it
[23:01] felixge has joined the channel
[23:01] felixge has joined the channel
[23:03] aryounce has joined the channel
[23:04] ollie has joined the channel
[23:05] kriszyp has joined the channel
[23:12] tilgovi has joined the channel
[23:16] creationix has joined the channel
[23:18] _ry: damn. i broke the build bot
[23:19] stalled_ has joined the channel
[23:21] JimBastard: thats it _ry you are fired
[23:21] JimBastard: please collect your things
[23:22] gwoo: haha
[23:47] aryounce: man, who scheduled JSconf so close to Drupalcon?
[23:48] aryounce: uncool
[23:48] rtomayko has joined the channel
[23:51] gf3 has joined the channel
[23:52] charlenopires has joined the channel
[23:57] technoweenie has joined the channel