[00:20] LordMetroid: Would it be insane to do a normal webpage using node.js? [00:21] Pilate: not if youre patient [00:21] LordMetroid: hmm, I am so lost in how to do my application [00:22] Pilate: theres a few djangoish routing systems out there already [00:25] mikeal has joined the channel [00:25] RayMorgan has joined the channel [00:27] taf2: i asked a little bit earlier but anyone looked into libdrizzle bindings? [00:50] jamiew has joined the channel [00:52] Booster has joined the channel [00:54] voodootikigod has joined the channel [00:58] jdzp has joined the channel [01:01] binary42 has joined the channel [01:13] jamiew_ has joined the channel [01:43] robrighter has joined the channel [01:43] robrighter has left the channel [01:48] tlrobinson1 has joined the channel [01:50] jamiew has joined the channel [01:56] tlrobinson1 has joined the channel [01:57] tlrobinson1 has joined the channel [02:02] taf2: any use vim? running node in !node foo.js scrambles my editor [02:06] Booster has joined the channel [02:18] geoff_blair has joined the channel [02:24] voodootikigod has joined the channel [02:41] jasondavies has joined the channel [02:43] kriszyp has joined the channel [02:44] kriskowal has joined the channel [03:00] jamiew_ has joined the channel [03:08] jdzp has joined the channel [03:08] mikeal has joined the channel [03:29] isaacs: ryah_away: url module (and uri deprecation) on my url-module branch [03:29] isaacs: oh, wait, nvm, i apparently broke a bunch of tests that relied all all sorts of url-parsing stuff coming from http.js... [03:29] isaacs: hehe [03:29] isaacs: tomorrow [03:30] jdzp has joined the channel [03:33] kriskowal: node and narwhal should converge on the uri module. there's no reason for it to be different. [03:33] kriskowal: in related news, we started with a uri module and just finished deprecating it in favor of url [03:34] isaacs: kriskowal: if it has a giant regex, that's a big drawback. [03:34] kriskowal: scratch that. we were as node is going, and going from where node is [03:34] isaacs: right [03:34] isaacs: i recall [03:35] kriskowal: why is the regex suboptimal? [03:35] charlenopires has joined the channel [03:35] kriskowal: i know it's hard to read; i had to break it up so i could maintain it. [03:35] isaacs: but the thing is, no one cares about any kinds of uris other than urls, and the regexp is slow to compile [03:36] kriskowal: let's consider those issues independently. [03:36] isaacs: also, it is hard to read, and i'm guessing it has a bug or two [03:36] kriskowal: the regex i think you're talking about is for url's [03:37] isaacs: the module i just wrote doesn't try to be 100% rfc-correct or enforcing, it just does what you expect. [03:37] kriskowal: is it slow to compile or slow to execute? has that been benchmarked? in comparison to an alternative? [03:37] kriskowal: uses indexOf scanning? [03:37] isaacs: in comparison to what's actually needed, which is just a handful of indexOf()s and substrs [03:37] kriskowal: i see. [03:37] isaacs: i'll do some rough and unscientific benchmarking tomorrow. [03:38] isaacs: but basically, the idea was to replace what's in node's http.js now, and that module has to absolutely fly [03:38] isaacs: anything your'e doing on any request must be blisteringly fast. [03:38] isaacs: s/any/every/ [03:38] isaacs: and actually, i ripped out any url parsing at all from http.js. it just gives you a string, you can parse it if you feel like [03:39] isaacs: just http.cat uses it, and that's not used too often anyway. does a full-buffer, etc. mostly just good for smallish things and debugging, not general-purpose fetching [03:45] kriskowal: you might want to pit the implementations against each other in tests [03:46] kriskowal: i'm sure correctness can be achieved both ways. i suspect that if you need a subset of what the regex impl provides, that a lighter regex will be faster than an implementation in terms of indexOf. [03:53] isaacs: maybe. [03:53] isaacs: k, i gotta run. have fun! [04:02] JimBastard has joined the channel [04:02] JimBastard: hello party people [04:03] inimino: hey JimBastard [04:22] scudco has joined the channel [04:25] RayMorgan has joined the channel [04:27] jed has joined the channel [04:34] JimBastard: hey inimino the express framework people were looking at your session code for doing sess managment [04:34] JimBastard: did you ever get on the github? [04:34] inimino: ah [04:34] inimino: no, there's a git repo though [04:35] inimino: just not on github [04:35] inimino: JimBastard: I will probably be doing another round of work on that in the next week or two in case they have any feature requests [04:41] JimBastard: there is a converation going on github :-\ [04:42] JimBastard: http://github.com/visionmedia/express/issues/#issue/60 [04:46] inimino: ah [05:03] sztanphet has joined the channel [06:01] BryanWB has joined the channel [06:02] BryanWB: I have a question about node.js concurrency that may be very obvious [06:03] BryanWB: if u create a web worker process from your main node.js process, can it run on other cpu cores/cpus on same machine or is it confined to the same cpu core of its parent process? [06:04] BryanWB: essentially, are any workers that your node.js process spawns confined to a single core? [06:15] mikeal has joined the channel [06:21] jamiew_ has joined the channel [06:25] micheil has joined the channel [07:10] jamiew_ has joined the channel [07:19] BryanWB: Are the web workers that your node.js process spawns confined to a single cpu core or can then run on any cpu core on any cpu on my machine? [08:29] qFox has joined the channel [09:08] gwoo_ has joined the channel [09:11] felixge has joined the channel [09:41] nsm has joined the channel [09:42] cloudhead has joined the channel [10:15] BryanWB has joined the channel [10:16] BryanWB: Are the web workers that a node.js process spawns confined to a single cpu core or can then run on any cpu core on any cpu on my machine? [10:19] felixge_ has joined the channel [10:27] sveisvei_a has joined the channel [10:35] tlrobinson has left the channel [10:35] felixge_: BryanWB: since they are independent processes your OS should assign them to any CPU as it sees fit (AFAIK) [10:36] BryanWB: felixge_, that's beautiful, though there could be locality issues there in terms of caching [10:37] felixge_: BryanWB: caching what? [10:37] BryanWB: felixge_, it would be best to stay on the same cpu, to take advantage of hardware memory buffers [10:38] BryanWB: felixge_, chances are that a new js process would refer similar libraries to its parent [10:38] felixge_: BryanWB: for some subsets of 'best', sure ;) [10:38] BryanWB: which would be cached by translation lookaside buffers and cache lines [10:39] felixge_: This will totally depend on what kind of stuff you are doing [10:39] felixge_: If your stuff does not benefit from running on more than 1 CPU, why spread it out into workers to begin with? [10:40] BryanWB: felixge_, true, I would only want to use more cpus/cores if I wanted to do something particularly computationally intensive in js [10:40] BryanWB: like crypto or a financial calculation [10:41] unom1 has joined the channel [10:41] BryanWB: it is this issue that has kept python from being used for really computationally intensive programming tasks [10:42] felixge_: yip [10:42] felixge_: anyway, web workers are not in node.js yet [10:42] BryanWB: but they are planned right? [10:43] felixge_: malte/cramforce has done an implementation, but its not official [10:43] felixge_: yes, they are planned [10:43] BryanWB: which is better than anything on python or ruby roadmaps [10:43] felixge_: not quite sure on the implementation details for the "real" ones so [10:43] felixge_: I think they'll use unix sockets to talk [10:43] felixge_: but I'm not sure if they'll spawn new processes or fork the main process [10:44] BryanWB: fork probably better in this case [10:44] BryanWB: s/this/many/ [10:45] teemow has joined the channel [10:54] mikeal has joined the channel [11:20] Biscuits has joined the channel [11:38] nsm has joined the channel [12:03] unom1 has joined the channel [12:06] nikhil_ has joined the channel [12:36] sveisvei has joined the channel [12:38] jasondavies has joined the channel [12:42] spoob has joined the channel [13:23] micheil has joined the channel [13:40] teemow has joined the channel [13:40] charlenopires has joined the channel [14:30] keeto has joined the channel [14:39] keeto has joined the channel [14:46] kriszyp has joined the channel [14:50] binary42 has joined the channel [15:03] alex-desktop has joined the channel [16:04] rictic has joined the channel [16:13] pjb3 has joined the channel [16:17] kriszyp has joined the channel [16:25] grantmichaels has joined the channel [16:39] botanicus_ has joined the channel [16:48] kriszyp_ has joined the channel [16:56] bryanl_ has joined the channel [17:13] Booster has joined the channel [17:16] softdrink has joined the channel [17:30] botanicus_ has joined the channel [17:36] Booster has joined the channel [17:40] jamiew has joined the channel [18:01] Booster has joined the channel [18:04] Booster has joined the channel [18:08] brandon_beacher has joined the channel [18:19] elliottcable: o7 all, what’s up? [18:29] robrighter has joined the channel [18:33] sudoer has joined the channel [18:57] Booster has joined the channel [19:05] gwoo has joined the channel [19:31] bryanl has joined the channel [19:38] bry has joined the channel [20:06] isaacs has joined the channel [20:12] mikeal has joined the channel [20:16] kriskowal has joined the channel [20:22] charlenopires has joined the channel [20:31] Booster has joined the channel [20:35] kriszyp has joined the channel [20:40] stephenlb has joined the channel [20:46] cloudhea has joined the channel [21:17] scudco has joined the channel [21:28] sztanphet has joined the channel [21:34] jed has joined the channel [21:51] Booster has joined the channel [21:53] isaacs: kriskowal: you around? [21:53] isaacs: http://github.com/isaacs/node-playground/blob/master/uri-vs-url.js [21:54] kriskowal: what's the score? [21:54] isaacs: http://gist.github.com/267693 [21:54] isaacs: now, it may be that the url module could go even faster. [21:54] isaacs: but it's like 20 vs 200 [21:55] isaacs: the parse failure is a small bug to fix, but the speed difference is significant. [21:55] kriskowal: good, good. API differences? [21:56] isaacs: plenty [21:56] isaacs: i wrote url to pretty much just provide the bare minimum, just like the browser does. [21:56] kriskowal: aye [21:56] isaacs: it does have "query" as well as "search", since it's easier to parse query strings if you don't have to pull off the ? first. [21:57] isaacs: but, it returns an object with href, host, hostname, auth, port, pathname, search, query, and hash [21:57] isaacs: oh, and protocol [21:57] charlenopires has joined the channel [21:57] kriskowal: i see. so it skips the recursive breakdown. [21:58] isaacs: right [21:58] kriskowal: that's fine. i'm assessing feasibility of interop [21:58] isaacs: it does have some duplicate stuff for hostname = auth+host+port, though [21:58] kriskowal: can you make it work in narwhal? [21:58] isaacs: and i'm also adding in my querystring module to node so that you can easily parse out what used to be on req.uri.params [21:59] isaacs: it's just javascript and string ops, so it should work fine, yeah. [21:59] isaacs: what's the main trunk for narwhal these days? still tlrobinson, or should i pull from 280north to get up to date? [21:59] kriskowal: 280north is the root now. [21:59] isaacs: keen [22:00] isaacs: oh, i should add resolve and format tests to this too [22:00] kriskowal: doesn't make much difference, really. depends on whose experiments you want to be subject to. [22:00] isaacs: hehe [22:00] kriskowal: that would be good. [22:00] isaacs: i think uri.resolveObject might be slightly faster, since it already has everything chopped up nice. [22:01] isaacs: directories split, etc. [22:02] isaacs: and it'd be unfair to test uri.resolve, since that would be subject to the parse time difference. [22:04] scudco has joined the channel [22:12] jed has joined the channel [22:19] cloudhead has joined the channel [22:28] rictic has joined the channel [22:40] grantmichaels has joined the channel [22:43] mattly has joined the channel [23:05] scudco has joined the channel [23:10] mattly has joined the channel [23:14] Booster has joined the channel [23:15] kriskowal has joined the channel [23:45] kriskowal has joined the channel [23:45] mattly has joined the channel [23:45] grantmichaels has joined the channel [23:45] rictic has joined the channel [23:45] jed has joined the channel [23:45] sztanpet has joined the channel [23:45] stephenlb has joined the channel [23:45] isaacs has joined the channel [23:45] bry has joined the channel [23:45] bryanl has joined the channel [23:45] gwoo has joined the channel [23:45] sudoer has joined the channel [23:45] brandon_beacher has joined the channel [23:45] pjb3 has joined the channel [23:45] alex-desktop has joined the channel [23:45] binary42 has joined the channel [23:45] teemow has joined the channel [23:45] jasondavies has joined the channel [23:45] sveisvei has joined the channel [23:45] Biscuits has joined the channel [23:45] ewdafa has joined the channel [23:45] Pilate has joined the channel [23:45] mjard has joined the channel [23:45] cadorn has joined the channel [23:45] halorgium has joined the channel [23:45] evilhackerdude has joined the channel [23:45] bengl has joined the channel [23:45] mikekelly has joined the channel [23:45] RJ2_ has joined the channel [23:45] eviltwin has joined the channel [23:45] Atmoz has joined the channel [23:45] jspiros has joined the channel [23:45] joshthecoder has joined the channel [23:45] rbranson_ has joined the channel [23:45] Yuffster has joined the channel [23:45] Vito` has joined the channel [23:45] jfoutz has joined the channel [23:45] hornbeck has joined the channel [23:45] wil has joined the channel [23:45] frodenius has joined the channel [23:45] jan____ has joined the channel [23:45] inimino has joined the channel [23:45] emyller has joined the channel [23:45] xantus has joined the channel [23:45] FoxFurry has joined the channel [23:45] elliottcable has joined the channel [23:45] CraigW has joined the channel [23:45] deanlandolt has joined the channel [23:45] erikcorry|away has joined the channel [23:45] beppu_ has joined the channel [23:45] skampler has joined the channel [23:45] orlandov has joined the channel [23:45] rudebwoy has joined the channel [23:45] jazzychad has joined the channel [23:45] onar has joined the channel [23:45] tlockney has joined the channel [23:45] ashb has joined the channel [23:45] martyn_ has joined the channel [23:45] rektide has joined the channel [23:45] [k2] has joined the channel [23:45] intellectronica has joined the channel [23:45] zimbatm has joined the channel [23:45] Sinjo has joined the channel [23:45] mediacoder has joined the channel [23:45] Sembiance has joined the channel [23:45] CIA-56 has joined the channel [23:45] ryah_away has joined the channel [23:54] mikeal has joined the channel [23:54] botanicus_ has joined the channel