[00:00] perezd: whats the preferred version to be compiling? [00:00] perezd: thats standard ubuntu 10.04 [00:00] bnoordhuis: perezd: yeah, i run that too [00:00] bnoordhuis: i take it you have libssl-dev installed? [00:00] mscdex: according to google it may have to do with server certificates [00:01] bnoordhuis: yeah, it's the error openssl raises if a cert doesn't validate [00:01] mscdex: :S [00:01] perezd: yeah we have that [00:01] bnoordhuis: but with a properly configured libssl you get an actual error message :) [00:01] chjj: latest version of the libssl-dev package seems to work for me [00:01] mscdex: is there a test for this? [00:01] mscdex: in node? [00:01] perezd: it happens to me randomly [00:01] bnoordhuis: mscdex: for what specifically? [00:02] perezd: I'm not even certain what causes it [00:02] perezd: exactly [00:02] mscdex: bnoordhuis: for checking that a useful error is given when the server cert isn't validated? [00:02] bnoordhuis: mscdex: no [00:02] mscdex: ah [00:03] Nuck has joined the channel [00:03] JoshC1 has joined the channel [00:04] bnoordhuis: perezd: the human-friendly error message is this: SSL routines:SSL3_GET_SERVER_CERTIFICATE:certificate verify failed [00:04] perezd: bnoordhuis: how does that happen really? [00:04] perezd: its a godaddy cert [00:04] perezd: is that why? :) [00:05] bnoordhuis: perezd: it's usually one of two things [00:05] bnoordhuis: perezd: either a misconfigured certificate chain [00:05] bnoordhuis: perezd: or an outdated root cert store [00:06] bnoordhuis: perezd: that's the certs in /etc/ssl/certs/ [00:06] CIA-65: node: 03Ryan Dahl 07master * rbceab40 10/ (deps/uv/src/win/process.c deps/uv/test/test-spawn.c): Upgrade libuv to c35548a - https://github.com/joyent/node/commit/bceab40b02f2da2ad282a0f649acf5ea251d2490 [00:07] bnoordhuis: perezd: try connecting directly with `openssl s_client -connect example.com:443` [00:07] Swizec has joined the channel [00:08] jmoyers has joined the channel [00:09] softdrink: how was i unaware of phantomjs's existence? [00:09] shoe has joined the channel [00:10] tbranyen: i dunno [00:10] tbranyen: softdrink: zombie.js is more friendly to Node.js users [00:10] kartmetal has joined the channel [00:11] ohtogo has joined the channel [00:12] softdrink: oh man…. the z-fighting in my little isometric-ish thing is really really bad in the latest chrome :( [00:13] ttpva has joined the channel [00:13] tmzt: z-fighting? [00:14] softdrink: http://dl.dropbox.com/u/543400/iso.html ← the flickering [00:14] dnunes: softdrink: There are some papers about z-sorting that can help you out. [00:14] softdrink: (flickers everywhere, but just really really bad in the latest chrome dev channel) [00:15] dnunes: btw: I don't get any flickering in Opera. [00:16] softdrink: does the 3d work in opera? i'm using webkit-specific vendor css [00:16] jbrokc: hello - can anyone help me understand what the Error( ) class is doing in the zombie.js source? it's coffeescript so be aware https://github.com/assaf/zombie/blob/master/src/zombie/bcat.coffee#L34 [00:16] telemachus has joined the channel [00:16] visudo has joined the channel [00:18] [[zz]] has joined the channel [00:18] bnoordhuis: jbrokc: is that a trick question? [00:19] Destos has joined the channel [00:19] CIA-65: node: 03Ryan Dahl 07master * ra44748b 10/ test/simple/test-child-process-env.js : win2k3: fix test-child-process-env - https://github.com/joyent/node/commit/a44748b3d7a53c710ee286c010e74d8aca117723 [00:20] [[zz]] has joined the channel [00:20] Aria has joined the channel [00:24] AvianFlu has joined the channel [00:25] Swizec has joined the channel [00:32] rfay has joined the channel [00:36] cafesofie has joined the channel [00:37] fermion has joined the channel [00:37] xerox has joined the channel [00:38] jacobolus has joined the channel [00:40] k1ttty has joined the channel [00:41] jamescarr has joined the channel [00:41] CIA-65: node: 03Ryan Dahl 07master * r2908f32 10/ (2 files in 2 dirs): win: fix test-child-process-exec-cwd - https://github.com/joyent/node/commit/2908f323e1cabd44329c8fcf67fd4b1fde41030b [00:42] cccaldas has joined the channel [00:43] jamescarr: content-after to generate a :? [00:43] jamescarr: ack, wrong channel [00:44] bbrandon has joined the channel [00:45] pylon__ has joined the channel [00:49] softdrink has joined the channel [00:50] levity_island has joined the channel [00:50] samsonjs has joined the channel [00:51] neoesque has joined the channel [00:51] levity_island has joined the channel [00:53] cafesofie has joined the channel [00:55] johnduhart has joined the channel [00:58] nphyx has joined the channel [01:01] seivan has joined the channel [01:02] monokrome has left the channel [01:05] dgathright has joined the channel [01:06] jbschw has joined the channel [01:08] ttpva has joined the channel [01:09] sirdancealot has joined the channel [01:09] gde33 has joined the channel [01:11] catb0t has joined the channel [01:11] nphyx has joined the channel [01:14] zackattack has joined the channel [01:14] ohtogo has joined the channel [01:15] skm has joined the channel [01:17] cl0udy has joined the channel [01:17] unlink has joined the channel [01:18] bayousoft has joined the channel [01:19] jbrokc has joined the channel [01:21] bentkus has joined the channel [01:22] zzo has joined the channel [01:23] Cromulent has joined the channel [01:24] krillr has joined the channel [01:24] krillr: ok, here's a question for you all [01:24] krillr: how the heck can you make an event handler know what its parent object is [01:24] krillr: here's the context: i have object A which has a function "eventHandler" [01:25] krillr: it wants to listen to object B's "test" emission, and do some action on object A AND object B [01:25] krillr: during emission "this" becomes the emitter, i lose all reference to the receiving object [01:25] krillr: how do i fix this? [01:25] AvianFlu: do you know about .call() ? [01:25] AvianFlu: and/or .apply() ? [01:25] krillr: i know of .call() but im unsure how that would apply here. and im unfamilair with .apply() [01:26] AvianFlu: so let's say you have objOne and objTwo [01:26] softdrink1 has joined the channel [01:27] AvianFlu: let's also say you have objOne.fnOne [01:27] akshatj has joined the channel [01:27] AvianFlu: you can do objTwo.call(objOne.fnOne, args) [01:27] CIA-65: node: 03Bert Belder 07master * r70bf121 10/ test/simple/test-child-process-cwd.js : Partial fix for test-child-process.cwd on windows - https://github.com/joyent/node/commit/70bf121f21c03d398a0d3c60e2088c29ff5a0b1b [01:27] AvianFlu: and the method will act like it's a member of the object you used .call on [01:27] abraxas has joined the channel [01:28] AvianFlu: so I figure name your event listeners, and you can just use .call as a fix [01:28] AvianFlu: unless I'm misunderstanding you [01:28] thepatr1ck: you can also do objOne.fnOne.bind(objOne) to get a function that has this = objOne [01:28] krillr: here, let me pastebin an example of what i wish to do [01:28] nphyx: an event handler function doesn't really have a parent object in a real sense, the "parent" object just has the function [01:29] ditesh|cassini has joined the channel [01:29] krillr: the "parent" object is a class in this case, i want to perform operations on that class [01:29] thepatr1ck: krillr: are you familiar with .bind() ? [01:29] krillr: indeed, but in a moment ill demonstrate how these solutions dont work. give me one sec [01:29] nphyx: On the class? Like, the prototype? You might be able to iterate through all the properties of all the objects in the global space and match them against the function or something like that but it would be horrible to do [01:30] CIA-65: node: 03Igor Zinkovsky 07master * r1c09cc1 10/ src/node_stdio_win32.cc : windows: set stdio streams to binary mode - https://github.com/joyent/node/commit/1c09cc1550af44eaf9d8ba6365d5d7911c41fce3 [01:31] bartt has joined the channel [01:31] nphyx: problem more or less is trying to map class-based inheritance to prototype based inheritance, it doesn't quite work 1:1. is there a way you can refactor your code so that this need goes away? [01:32] krillr: http://paste.pocoo.org/show/450940/ [01:32] krillr: here's what im trying to achieve [01:33] krillr: er, oens ec [01:33] jmoyers: a.on('test', b.eventHandler); -- so in this case [01:33] jmoyers: what do you want `this` to refer to? [01:34] krillr: http://paste.pocoo.org/show/450941/ fixed [01:34] krillr: jmoyers, "this" should refer to b [01:34] Dmwuw has joined the channel [01:35] nphyx: you could put your callbacks in a closure with a self-referencing var, maybe? e.g. b.callback = function() {var firstClass = this; var foo = function() {... callback stuff...}; return foo;} [01:36] nphyx: then inside foo firstClass should refer to b, or whatever. sorry to swap your conventions around. [01:36] krillr: ACTION screams [01:36] krillr: lol [01:37] krillr: trouble is [01:37] krillr: i dont want to have to know who the emitter or listener is [01:37] krillr: what class they're from [01:37] krillr: it should "just work" [01:37] jmoyers: nah nah [01:38] jmoyers: here [01:38] jmoyers: b.eventHandler.bind(b) [01:38] jmoyers: simple [01:38] jmoyers: no retarded closures needed [01:38] nphyx: but then you lose a, jmoyers [01:38] nphyx: retarded closures haha : ) [01:38] jerrysv has joined the channel [01:39] jmoyers: he said `this` should refer to b [01:39] jmoyers: wheres the need for ref to a? [01:39] Swimming_bird has joined the channel [01:40] jmoyers: if you need to know who emitted [01:40] nphyx: it wants to listen to object B's "test" emission, and do some action on object A AND object B [01:40] nphyx: during emission "this" becomes the emitter, i lose all reference to the receiving object [01:40] jmoyers: i'd make the sig function(emitter, payload){} [01:40] jmoyers: yes, so with bind, this is preserved as the listener [01:40] jmoyers: problem solved [01:41] mbrevoort has joined the channel [01:41] krillr: can you reply to my paste with a working example? [01:41] jmoyers: no? [01:41] nphyx: actually here's the problem: a.on('test', b.eventHandler); b.eventHandler is losing its context [01:41] krillr: im unsure how to apply bind in this manner [01:41] krillr: right [01:41] zackattack has joined the channel [01:42] Anton_ has joined the channel [01:42] jmoyers: yes, yes it is [01:42] nphyx: a.on('test', function() {b.eventHandler(this)}) is what you want [01:42] jmoyers: and bind preserves the context [01:42] AvianFlu: you're both right [01:42] nphyx: b.eventHandler's sig becomes function eventHandler(emitter) { ... this.doStuffTo(emitter)} [01:42] krillr: whoa [01:42] krillr: nphyx, thats perfect [01:43] jmoyers: replied [01:43] jakehow has joined the channel [01:43] kriszyp has joined the channel [01:43] jbschw has joined the channel [01:44] nphyx: it's hard to remember coming from c-style languages but when you pass a function by reference it always loses its context [01:45] JumpMast3r has joined the channel [01:46] bingomanatee has left the channel [01:46] jhurliman has joined the channel [01:46] devongovett has joined the channel [01:47] thinkt4nk has joined the channel [01:47] brolin has joined the channel [01:50] jbrokc has joined the channel [01:50] ank has joined the channel [01:52] d0k has joined the channel [01:53] odie5533 has joined the channel [01:53] stalled has joined the channel [01:54] jamescarr: bout to make a sick sick backbone based application [01:54] jamescarr: wish me luck amigos [01:55] merong has joined the channel [01:55] thinkt4nk: heh, luck [01:55] thinkt4nk: you doin' it in one sitting or something? [01:56] softdrink has joined the channel [01:57] merong: andrew [01:57] merong: mosco [01:57] merong: hinek [01:57] merong: mike [01:58] jamescarr: thinkt4nk, YES [01:58] jamescarr: I need to be live in 42 hours :) [01:58] merong has left the channel [01:58] jamescarr: back end is really done... it's the front end ;) [01:58] thinkt4nk: hah, damn! [01:58] thinkt4nk: hold on to your hat! [01:58] thinkt4nk: lol [01:58] thinkt4nk: I hope you have enough caffeine, dude. [01:59] nphyx: 42 hours damn man you do need luck :) [01:59] thinkt4nk: and lots of unit tests lol [02:00] krillr: woot, it worked [02:00] krillr: thanks guys [02:03] odie5533: I am having trouble installing npm and am getting permission errors. The only thing I found was to chown the entire /usr/local to my user, but this seems like a bad idea. Is there anyway to install it without screwing with my system's permissions? [02:03] jensn has joined the channel [02:04] nphyx: chowning /usr/local is the typical solution [02:04] nphyx: you can install it to your home folder if that works [02:04] nphyx: just set prefix to ~/node/ or whatever during compile [02:05] nphyx: and add that to your PATH for ease of use [02:06] odie5533: is NPM written in Nodejs? [02:06] nphyx: there's an article, i think on howtonode.org, that goes over the various workarounds to that problem [02:06] Dreamer3 has joined the channel [02:07] cjm has joined the channel [02:07] Sorella has joined the channel [02:09] catb0t has joined the channel [02:10] nphyx: oh! npm, not node, sorry. yeah npm is written in node [02:11] odie5533: bootstrapping node. Now there's a project. [02:11] tmzt: it's not an issue on recent versions of node [02:11] nphyx: you *may* be able to run npm out of a local dir with node ~/path/to/npm.js [02:11] zaach has joined the channel [02:12] tmzt: npm is supposed to run in the current directory or the root of your poject [02:12] tmzt: project [02:12] tk has joined the channel [02:14] jppjcnva has joined the channel [02:16] jppjcnva has left the channel [02:16] jppjcnva has joined the channel [02:17] balaa has joined the channel [02:18] Nuck: I overuse the require()() pattern. [02:19] Nuck: But damnit, it's awesome for sharing variables with modular code. [02:19] Nuck: require('./batches/accounts.js')(server, db); [02:19] Nuck: Look at that sexy beast :P [02:19] thinkt4nk: ACTION likes [02:20] jesusabdullah: It appears to be a beast of the non-sexy variety! [02:20] Nuck: I personally disagree [02:20] jesusabdullah: I prefer doAmazingThing(require('middlewarez')); [02:20] rickharrison has joined the channel [02:20] Nuck: I can encapsulate code while still allowing a degree of connection [02:20] Nuck: Ah, middlewaring it up [02:20] wearebugs has joined the channel [02:20] monokrome has joined the channel [02:21] jesusabdullah: indeed [02:21] Nuck: I prefer to use the curry-style pattern personally, it fits my general use-cases better. [02:21] jesusabdullah: Actually, I'm not much of a hater for, well [02:21] jesusabdullah: there aren't too many of those sorta things that I actually dislike [02:21] dherman has joined the channel [02:21] jesusabdullah: I can't think of any right now [02:21] jesusabdullah: though I lurve method chaining [02:21] jbschw has left the channel [02:21] WarriorSl has joined the channel [02:22] Nuck: jesusabdullah: My favorite line is this one: server.use(require('./lib/uploads.js')('/tmp/uploads/')); [02:22] jesusabdullah: Me gusta. [02:22] Nuck: require() returns a function which I call immediately with the config, then I pass the function it returns to another function. [02:22] Nuck: It's FUNCTIONCEPTION [02:23] apanda has joined the channel [02:24] samsonjs has joined the channel [02:24] yuroyoro has joined the channel [02:24] jamescarr: tj [02:24] NickABusey has joined the channel [02:24] jamescarr: tjholowaychuk, is there a simple way to "alias" routes? [02:25] jamescarr: like routes that are under a path like /foo/bar? [02:25] jamescarr: (dumb question I know) [02:25] tjholowaychuk: jamescarr: just pass the same callback to a few route definitions [02:26] stalled has joined the channel [02:28] Nuck: tjholowaychuk: You should make an Express-based load balancer/router thingy [02:28] Nuck: So I never hafta use Nginx lol [02:28] pen has joined the channel [02:31] hippich has joined the channel [02:31] hippich has joined the channel [02:31] clifton_: you can do it all with node-http-proxy, actually [02:31] Nuck: Nginx is probably more efficient lol [02:31] clifton: yes, it is [02:31] Nuck: There are very few people here who could manage to beat that kinda efficiency. [02:32] Nuck: (folks like tj and izaac) [02:32] clifton: well, its not going to be beaten by something written in javascript [02:33] losing has joined the channel [02:33] Nuck: Every program has issues [02:33] clifton: honestly node-http-proxy is probably find for load balancing for all but insane scenarios [02:33] clifton: nodejitsu.com uses it in production [02:33] Me1000 has joined the channel [02:33] jamescarr: tjholowaychuk, btw, epic win with stylus [02:33] tjholowaychuk: jamescarr thanks man [02:33] jamescarr: I have a rather large loan management app written in express+jade+stylus [02:33] Nuck: I want to eventually grow to a community of hundreds of thousands of members. [02:34] jamescarr: doing a bit of tweaking to it now ;) [02:34] clifton: it would be fine for that. id just be hesitant to have it serve static assets and do other stuff that can be done without hitting the node stack [02:34] Nuck: My biggest problem is wondering whether my load balancers will have trouble whenever they fail over [02:35] Nuck: Because deviantART has that problem [02:35] cha0s has joined the channel [02:35] ns5 has joined the channel [02:35] Nuck: Every time the load balancer in front of their chat server, and all the sockets disconnect and the server goes wonky whenever it fails over. [02:35] Nuck: And I'm gonna have a chat server, and I don't want it to have that kinda problem. [02:36] cl0udy_ has joined the channel [02:36] `3rdEden has joined the channel [02:37] jamescarr: cd #zsh [02:37] ns5: User sends a message to node.js running on server, node.js needs to forward this message to another daemon written by C, this C daemon does all the calculation to decide which channels this message should be sent to. How to implement the communication between node.js and C daemon? [02:37] Aria: If you control the client, you can handle that problem in the client. [02:37] Aria: Transferring live sockets that dumb clients use is harder. [02:38] Nuck: Aria: Is that for me or ns5? [02:38] Aria: you, Nuck. [02:38] jacobolus has joined the channel [02:38] Nuck: ah [02:38] Nuck: Aria: Can socket.io do that stuff for me? [02:38] pifantastic: #lazyweb what redis client are the cool kids using? [02:39] Aria: Most of it, perhaps not all. Might need some custom reconnect logic. [02:39] pifantastic: https://github.com/mranney/node_redis/ has the most GH followers [02:39] Nuck: Aria: It's not that the server changes [02:39] Nuck: It's the same chat server [02:39] tbranyen: it used in connect-redis too [02:39] Aria: ns5: Nothing special -- send the message using some sort of delimited socket stream. Keep it simple, unless some part is already set in stone in some way. [02:40] pifantastic: tbranyen: node_redis is? [02:41] emmanuel has joined the channel [02:43] boogyman has joined the channel [02:43] wookiehangover has joined the channel [02:45] tbranyen: pifantastic: yea [02:45] ngs has joined the channel [02:45] bayousoft has joined the channel [02:46] ryanfitz has joined the channel [02:46] odie5533: I am trying to do some simple web scraping. Is there a cookie jar-like thing for nodejs? [02:46] pifantastic: tbranyen: sweet, thanks [02:47] espacia: odie5533: you're looking for a particular library for scraping web content? [02:47] espacia: odie5533: what type of content? [02:47] odie5533: espacia: Any library. HTML content [02:48] Aria: Zombie's not bad. [02:48] odie5533: But I want it to be able to perform simple login/cookie procedures. I've done it in Python using the Twisted Python library. [02:49] tbranyen: Zombie is really good has a few issues tho [02:49] tbranyen: i think it might help to update jsdom manually [02:49] ank has joined the channel [02:49] odie5533: Zombie looks like it's for unit tests more than anything [02:49] tbranyen: nah you just use unit tests with it [02:49] mape: ryah: http://www.youtube.com/watch?v=1eYaq358LRQ [02:49] tbranyen: you can do whatever [02:50] hippich has joined the channel [02:51] espacia: odie5533: try these libs: https://github.com/arrix/node-readability, https://github.com/kirbysayshi/node-htmlparser [02:51] CIA-65: libuv: 03Bert Belder 07spawn-err-async * ree68c09 10/ (include/uv-win.h src/win/process.c): Windows: report uv_spawn() failure asynchronously - https://github.com/joyent/libuv/commit/ee68c099071cd1fb9b14b8b0777f0101afd08f1f [02:52] yhahn has joined the channel [02:54] aurojit has joined the channel [02:54] seivan has joined the channel [02:55] ghost has joined the channel [02:57] gazumps has joined the channel [02:57] chjj has joined the channel [03:02] cl0udy has joined the channel [03:02] cafesofie has joined the channel [03:03] DEVX__ has joined the channel [03:03] Drakonite has joined the channel [03:03] CIA-65: libuv: 03Bert Belder 07spawn-err-async * r37cca1c 10/ (include/uv-win.h src/win/process.c): Windows: report uv_spawn() failure asynchronously - https://github.com/joyent/libuv/commit/37cca1c7e35666b10f9757a3392485307d496135 [03:07] Wizek has joined the channel [03:07] clifton has joined the channel [03:09] balaa has joined the channel [03:11] CIA-65: libuv: 03Igor Zinkovsky 07master * r6d940a3 10/ (6 files in 4 dirs): windows: remove dependency on rpcrt4 and ole32 libs. fixes https://github.com/joyent/libuv/issues/118 - https://github.com/joyent/libuv/commit/6d940a3afcbd4280801f9adfb72c821c4edbccf9 [03:11] hacksparrow has joined the channel [03:12] hippich has joined the channel [03:12] hippich has joined the channel [03:12] hacksparrow has joined the channel [03:13] JakeyChan has joined the channel [03:15] slifty has joined the channel [03:15] gazumps has joined the channel [03:15] mikeal has joined the channel [03:17] zanes has joined the channel [03:17] AvianFlu has joined the channel [03:17] juli0w has joined the channel [03:22] MikhX has joined the channel [03:22] towski has joined the channel [03:24] maru_cc_ has joined the channel [03:25] CIA-65: libuv: 03Bert Belder 07master * r2e9a743 10/ (include/uv-win.h src/win/process.c): Windows: report uv_spawn() failure asynchronously - https://github.com/joyent/libuv/commit/2e9a743da08d96ac7861400a316e3a9d3067979f [03:25] mykul has joined the channel [03:25] levi501d has joined the channel [03:27] CIA-65: node: 03Ryan Dahl 07master * r89b14fc 10/ (8 files in 5 dirs): Upgrade libuv to 2e9a743 - https://github.com/joyent/node/commit/89b14fc0ff4da52ec112e0af4d8e7d87bfe5d637 [03:28] Will| has joined the channel [03:28] Leonidas has joined the channel [03:33] domo has joined the channel [03:34] domo: hey guys - whats the best way to open a template file, get the contents and give to ejs for parsing? [03:34] domo: the EJS docs don't cover that and I'm thinking just read the file using node js' built in fs lib? [03:34] domo: I'm not using express - I don't need it. connect will do just fine [03:35] apanda has joined the channel [03:35] ericmuyse has joined the channel [03:37] tristanseifert has joined the channel [03:37] jerrysv: domo: fs.readFile ? [03:38] domo: jerrysv: so, I'm guessing to keep 'dry', most people write a util function to handle the 'rendering'? [03:38] Druid has joined the channel [03:38] domo: I guess Im asking if there is anything built in that reads file and parses that im missing :) [03:38] jerrysv: domo: i don't use connect -- i was just advising on getting a file into memory :) [03:42] harthur has joined the channel [03:43] vns has joined the channel [03:44] CIA-65: node: 03Igor Zinkovsky 07master * ra6e0a91 10/ wscript : windows: remove dependency on rpcrt4 and ole32 - https://github.com/joyent/node/commit/a6e0a91a70ed450166a2066fe2a7d6a0d3969ca9 [03:44] bpwnr has joined the channel [03:46] __directory: mongo pretty decent with single server now that there's journaling? [03:49] slifty has joined the channel [03:50] theCode has joined the channel [03:51] dgathright has joined the channel [03:53] jmoyers has joined the channel [03:53] gerard0 has joined the channel [03:54] mikeal has joined the channel [03:54] themiddleman has joined the channel [03:55] dnunes has left the channel [03:56] hassox has joined the channel [03:59] monokrome has joined the channel [04:00] andrewfff has joined the channel [04:00] mikeal has joined the channel [04:02] z8000 has joined the channel [04:03] cafesofie has joined the channel [04:04] ericmuyser has joined the channel [04:05] zackattack has joined the channel [04:05] wookiehangover has joined the channel [04:05] piscisaureus has joined the channel [04:12] apanda has joined the channel [04:13] ank has joined the channel [04:14] meso_ has joined the channel [04:14] apanda has joined the channel [04:15] meso__ has joined the channel [04:16] yhahn has left the channel [04:16] idefine has joined the channel [04:17] ChrisPartridge has joined the channel [04:17] aguynamedben has joined the channel [04:20] slajax1 has joined the channel [04:21] aguynamedben_ has joined the channel [04:22] zeade has joined the channel [04:22] theCode has joined the channel [04:22] rudolfrck has joined the channel [04:23] aguynamedben has joined the channel [04:25] Corren_ has joined the channel [04:25] fairwinds has joined the channel [04:26] rurufufuss has joined the channel [04:26] JakeyChan has joined the channel [04:27] jMerliN has joined the channel [04:27] boaz has joined the channel [04:31] aguynamedben_ has joined the channel [04:32] jesswa has joined the channel [04:34] jesswa: anybody here familiar with bcrypt and/or mongoose-auth? I'm getting errors from the bcrypt when trying to run the mongoose-auth examples... [04:34] brianc has joined the channel [04:34] al3xnull has joined the channel [04:35] gg777 has joined the channel [04:35] ChrisPartridge: jesswa: what errors? [04:36] luckysmack has joined the channel [04:37] jesswa: ChrisPartridge http://pastebin.com/PT15BDwg [04:38] jesusabdullah: jesswa: Were you the one looking to bcrypt chats? [04:38] jesusabdullah: like a week ago? [04:39] jesswa: jesusabdullah: I don't think so - just using it as part of everyauth/mongoose-auth [04:39] jesusabdullah: ah [04:39] jesusabdullah: INTERESTING [04:40] jesusabdullah: What's an everyauth, anyway? :/ [04:40] jesusabdullah: I just want auth to not suck <_< [04:40] __directory: im wanting to know if mongo 1.8 is reliable single server with --journaling [04:40] __directory: make authnosuck then [04:40] jesusabdullah: __directory: Did you follow me on twitter? [04:40] jesusabdullah: ACTION crosses his arms [04:40] sw8sw8 has joined the channel [04:40] jesusabdullah: ACTION puts his hands on his hips [04:40] Aria: __directory: defined reliable? [04:41] __directory: I normally use Postres - so of that level [04:41] ChrisPartridge: jesswa: did bcrypt compile correctly? [04:41] __directory: jesusabdullah: yes I did [04:41] jesswa: jesusabdullahL everyauth is just an auth module for express [04:41] jesusabdullah: hah! [04:41] jesswa: ChrisPartridge: everything looked fine, got two warning/info messages: [04:41] jesswa: Checking for node path: not found [04:42] jesswa: and then "[4/4] cxx_link: build/default/src/blowfish_1.o build/default/src/bcrypt_1.o build/default/src/bcrypt_node_1.o -> build/default/bcrypt_lib.node" [04:42] jesusabdullah: Aria: My robot name is like ee cummings style cat names---nobody gets to know my (secret) (real) robot name! [04:42] jesswa: (when installing bcrypt) [04:42] jesswa: installed via npm [04:42] jesswa: I can paste the full output to another pastebin if you want [04:42] __directory: Aria: so of Postgres caliber [04:43] Aria: __directory: It's pretty reasonable, yeah. It's not ACID or MVCC, so it's a different use case, but … yeah. [04:43] andrewfff has joined the channel [04:43] __directory: this is for an internal customer management thing for 5 people [04:44] __directory: I only have 10,600 customer records [04:44] tbranyen: yeah well i only a million [04:44] tbranyen: take that [04:45] __directory: we're a small place [04:45] __directory: we have a lot of guns and knives [04:45] pifantastic_ has joined the channel [04:45] ank has joined the channel [04:47] Aria: That seems perfectly reasonable. [04:48] __directory: I'll try it with mongo then. Ill replicate too [04:48] aguynamedben_ has joined the channel [04:50] ryanfitz has joined the channel [04:50] brianc: __directory: why not use postgres? [04:51] kriszyp has joined the channel [04:52] CIA-65: node: 03Ryan Dahl 07master * r6d5218b 10/ (4 files in 3 dirs): [04:52] CIA-65: node: Merge branch 'v0.4' [04:52] CIA-65: node: Conflicts: [04:52] CIA-65: node: doc/api/crypto.markdown [04:52] CIA-65: node: doc/api/modules.markdown [04:52] CIA-65: node: src/platform_win32.cc - https://github.com/joyent/node/commit/6d5218bc7d344b0f4e0a0cb50fdd25052946cc4f [04:52] __directory: brianc: I dont have a good reason not to. I use it for everything else. [04:53] brianc: __directory: just consider the next person who comes into the job having to learn postgres _and_ mongodb [04:54] __directory: brianc: are people trying to write orms on top of node-postgres or just abstracting client.query calls or? [04:54] jesswa has joined the channel [04:55] brianc: __directory: few different orms [04:55] brianc: ormnomnom and FastLeGs I believe [04:56] brianc: i don't really follow on it too too much. I really try to keep my OS project count low so I can fix bugs as fast as possible on node-postgres [04:56] brianc: would rather provide one solid library than 10 wishy washy ones [04:56] dshaw_ has joined the channel [04:56] brianc: __directory: i mostly just use client.query. I been working on a sql generator for something like: users.select(users.firstName).where(users.firstName.like('blah')) but i always hit a wall with ORMs and end up missing real SQL [04:56] yozgrahame has joined the channel [04:56] zackattack has joined the channel [04:57] __directory: ORMS are weird [04:58] brianc: __directory: yeah, the 'easy case' works nicely, but once you need an average of length of audio grouped by top level hierarchy account (recursive children) can't really do that w/ orms anymore [04:58] jesswa has joined the channel [04:59] brianc: __directory: anyways...not really trying to sell node-postgres as much as just recommending you consider the full cost and advantages of introducing a new dependency to your companies infrastructure. [04:59] brianc: *company's [04:59] brianc: and now i gotta go to bed [04:59] brianc: good luck either way. :) [05:00] __directory: I've owned it for 20 years and am pretty good at introducing costs [05:00] __directory: ya thanks brianc [05:10] jmoyers has joined the channel [05:12] jesswa1 has joined the channel [05:13] slifty: night night everyone! [05:16] __directory: night!!!!! [05:18] jesswa has joined the channel [05:18] evlapix has joined the channel [05:19] evlapix: Had anybody ever had a "throw new Error" not throw an error? [05:20] losing has joined the channel [05:21] samsonjs has joined the channel [05:22] rchavik has joined the channel [05:22] seivan has joined the channel [05:26] ryah: test please http://nodejs.org/dist/v0.5.3/node-v0.5.3.tar.gz [05:27] jesswa has joined the channel [05:28] jesswa1 has joined the channel [05:29] luckysmack: ryah, to see if the download works? or the whole shebang? [05:29] mikeal has joined the channel [05:32] smathy has joined the channel [05:33] jesusabdullah: HOLY SHIT I just learned that I can edit other peoples' comments on my github issues [05:33] luckysmack: lol [05:33] Lorentz: Now now, don't use it for evil [05:33] Lorentz: Even if it's funny [05:34] captain_morgan has joined the channel [05:34] jesusabdullah: I know :( [05:34] jesusabdullah: but the troll in me is, well [05:34] CIA-65: node: 03Ryan Dahl 07master * r3ce6712 10/ (wscript src/node_child_process_win32.cc): Remove node_child_process_win32.cc - https://github.com/joyent/node/commit/3ce671284b61c34767994d94548a27836e706735 [05:34] jesusabdullah: I'm sure you can imagine [05:34] levity_island has joined the channel [05:35] losing has joined the channel [05:35] jesusabdullah: I wouldn't be mean, since the person whose comment I noticed the "edit" button on gave me a pretty good pointer [05:35] __directory: heh [05:37] TooTallNate has joined the channel [05:38] jesswa has joined the channel [05:39] newy has joined the channel [05:39] Nexxy has joined the channel [05:39] Nexxy has joined the channel [05:40] losing_ has joined the channel [05:43] killfill has joined the channel [05:44] __directory: could really use a tug off the ole' crack pipe right about now [05:45] mikeal has joined the channel [05:45] losing has joined the channel [05:46] jesswa has joined the channel [05:46] loob2 has joined the channel [05:46] Jippi has joined the channel [05:46] TheJH has joined the channel [05:49] Nexxy: __directory, how offensive! [05:50] admc has joined the channel [05:51] CIA-65: node: 03Ryan Dahl 07master * r8ace421 10/ src/node_extensions.h : Complete removal of node_child_process_win32.cc - https://github.com/joyent/node/commit/8ace421bfda284f600edb5aeaa525b7817078922 [05:51] JoshC1 has joined the channel [05:53] maru_cc_ has joined the channel [05:54] __directory: :( [05:54] mscdex: you tell them Nexxy [05:55] losing_ has joined the channel [05:58] SamuraiJack has joined the channel [05:59] stephank has joined the channel [06:02] chjj: ah, i see github decided to go the oversized footer route [06:02] chjj: shame [06:04] smathy: Heh :) [06:06] Jakwac has joined the channel [06:09] levity_island: yeah but that octobiwan mug is win [06:10] jesswa1 has joined the channel [06:10] smathy: CSS makes fools of the best of us. [06:11] jamescarr: just use stylus [06:11] jamescarr: stylus or less or sass ;) [06:11] meso_ has joined the channel [06:12] smathy: Ok, so more that CSS combined with DOM selector rules makes fools of the best of us. [06:14] apanda has joined the channel [06:15] Wizek has joined the channel [06:17] _bat has joined the channel [06:23] mhausenblas has joined the channel [06:25] shinuza has joined the channel [06:26] svenlito has joined the channel [06:27] kartmetal_ has joined the channel [06:27] kartmetal_ has joined the channel [06:27] e6nian has joined the channel [06:28] der| has joined the channel [06:28] der|: Hi there! [06:29] der|: I've been looking for a decent templating system for Node, and none suit my workflow. I've created something interesting just now :) [06:30] ckknight: oh? [06:30] __directory: ! [06:32] der|: yeah [06:32] der|: I come from PHP, and I don't quite like the current methods, so I made node behave exactly like PHP & Apache [06:32] der|: you can mix real code & presentation [06:33] tbranyen: wat [06:33] der|: for example: http://paste.pocoo.org/show/451040/ [06:33] tbranyen: der|: you are aware of the other template engines that do that right? [06:33] tbranyen: there are tons of those [06:33] _bat: I kinda like dust.js [06:33] der|: tbranyen: yup [06:33] simenbrekken has joined the channel [06:33] tbranyen: like almost identical to that [06:33] der|: Here's the output: http://paste.pocoo.org/show/451044/ [06:33] tbranyen: why make your own? [06:33] tbranyen: for fun? [06:34] der|: not really, can you execute real js code inside the templates ? [06:34] tbranyen: yes. [06:34] der|: with dust.js ? [06:34] der|: let's see heh [06:34] Nuck: Hurrah! Got a gitolite set up without a single kink this time. [06:35] tbranyen: underscore templates, ejs, that new one jon neal made [06:35] der|: nope, dust.js does not do what mine does heh [06:35] der|: you have to use {} for vars [06:35] der|: this one behaves exactly like php [06:35] tbranyen: well thats a deal breaker [06:35] tbranyen: except the fact that its not php [06:35] der|: I mean, if you come from php, it will be painless [06:36] der|: it's still js, but you use similar escapes. In this case: [06:36] tbranyen: nearly all template engines that don't suck allow you to customize those [06:36] __directory: hmm [06:37] Nuck: EJS [06:37] der|: I'm not talking about customization per se, the open/close tags are irrelevant [06:37] Nuck: Great for PHP [06:37] Nuck: Well, ex-PHP users [06:37] der|: Nuck: exactly :) [06:37] Nuck: I love EJS because it provides all the simplicity of PHP combined with the awesomeness of JS [06:38] der|: I think I just did the same thing as EJS then lol [06:38] booo has joined the channel [06:38] blkcat: yup. [06:38] __directory: c'mon Nuck [06:39] der|: haha lol [06:39] der|: well, mine is more optimal [06:39] Nuck: __directory: I'm gonna take a wild guess and say you're about to make a Ruby joke, huh? [06:39] Nuck: Howso? [06:39] __directory: ruby joke is on you [06:39] Nuck: __directory: Wow, that's a first. [06:40] blkcat: man, weird to think it's been almost two years since _why vanished [06:40] groom has joined the channel [06:40] der|: EJS is 505 lines of code. Mine is 123 w/comments [06:40] apanda has joined the channel [06:40] Nuck: blkcat: Never knew the guy. [06:40] Nuck: Never learned Ruby. [06:40] Nuck: Despise it. [06:40] blkcat: Nuck: i never accused you of either. :) [06:41] Nuck: der|: EJS for the client or server? [06:41] blkcat: never used ruby myself, but i still miss _why. [06:41] __directory: you were just the other day talking Ruby, setting up Redmine [06:41] __directory: now you claim to have never used it? [06:41] Nuck: It also provides some things like <%=variable %> [06:41] Nuck: __directory: Never used the language [06:41] __directory: ehh..wow. [06:41] Nuck: Have used programs built on it [06:41] der|: Nuck: I'm not quite sure... it only has ejs_production.js, ejs.js & view.js [06:41] V1 has joined the channel [06:41] mhauri has joined the channel [06:41] Nuck: Sorta like how I've never used C++, but I have used Node. [06:41] der|: I assume it's ejs.js [06:42] SubStack: ruby is pretty neat [06:42] Nuck: SubStack: STFU, be our _why. [06:42] balupton has joined the channel [06:42] SubStack: Nuck: you want me to delete all my repos then? >:D [06:42] Nuck: SubStack: lolno [06:43] levi501d has joined the channel [06:43] Nuck: ACTION goes to clone all of SubStack's repos in preparation for the Subpocalypse [06:43] groom has joined the channel [06:43] __directory: could make a nuck bot [06:43] __directory: like the izs thing [06:43] gozala has joined the channel [06:43] _bat: Nuck: I haven't seen any .npmignore files in SubStack's repos, so I think we're good [06:43] thalll has joined the channel [06:43] Nuck: __directory: Every response would be "fuck ruby", "fuck CoffeeScript" or some variant thereof. [06:44] der|: hahaha [06:44] Nuck: So, if I *do* get into Y Combinator (crossing my fingers) I guess I'll be moving to the bay area lol [06:45] SubStack: Nuck: I wouldn't plan on it [06:45] SubStack: come to the bay anyways [06:46] SubStack: it's a fun place [06:46] Nuck: SubStack: I hope to. I just need some way to fund it, and YC is ideal. [06:46] _bat: down by the bay, where the watermelon grows [06:47] mhausenblas_ has joined the channel [06:47] Nuck: I think I *might* have a chance at YC. We've got a damn good team going, and we've grown so far beyond "generic deviantART clone" (not telling how though) [06:47] balupton has joined the channel [06:47] bbrandon: Nuck, tell us about your startup? [06:48] Nuck: bbrandon: eh? It's an art site, like deviantART, but we're being stealthy about the details >_> [06:48] bbrandon: Sounds exiting! [06:48] bbrandon: Are you like Art.sy? [06:49] Nuck: ACTION looks it up [06:49] __directory: im doing facebook, but you can have your dog too. [06:49] __directory: what's the cost of living like in the Bay? [06:50] Nuck: tbranyen: Not really, we're gonna be focusing less on gallery art, and more on the average "starving artists", like deviantART has. [06:50] ambroff has joined the channel [06:50] Nuck: Our primary competitor will likely end up being dA, though we have nothing but the utmost respect for them. [06:51] Nuck: err, bbrandon [06:51] Nuck: ACTION derped there [06:51] Nuck: Damn all you *bra* names [06:51] Nuck: Wait [06:51] Nuck: lolbra [06:52] __directory: that's right, brah [06:53] _bat: I'm surprised more people don't put _ in front of their names [06:53] _nuck: lol [06:53] __directory: i did it twice, just to prove a point [06:53] ___nuck: PWN'D [06:54] __directory: doh [06:54] thalll has joined the channel [06:54] _bat: hmm…[ is showing up higher than _ for me [06:54] ___nuck: same [06:55] cjm has joined the channel [06:58] hassox has joined the channel [07:00] ttpva has joined the channel [07:01] bbrandon: NUCK! [07:03] __tosh has joined the channel [07:03] Ramosa has joined the channel [07:06] cosmincx has joined the channel [07:07] fangel has joined the channel [07:07] socketio\test\63 has joined the channel [07:07] socketio\test\52 has joined the channel [07:07] mrtrosen has joined the channel [07:08] Elems has joined the channel [07:08] CIA-65: libuv: 03Ryan Dahl 07master * r22ec2db 10/ (config-unix.mk src/uv-unix.c): solaris fixes - https://github.com/joyent/libuv/commit/22ec2db4134953e1fffedb7fd1b6fd78d12e5293 [07:08] CIA-65: libuv: 03Ryan Dahl 07master * rb2ed24d 10/ (include/uv-unix.h src/uv-unix.c): [07:08] CIA-65: libuv: unix: Remove flock code [07:08] CIA-65: libuv: solaris doesn't have flock - https://github.com/joyent/libuv/commit/b2ed24d86ecc561135edf566878b43d93e364d44 [07:08] Ramosa: Something very odd.. I couldn't install Coffee-script by typing npm install coffee-script when logged into Debian by VirtualBox|Terminal. Got Error: ENOENT. No such folder coffee-script.. but when I do the same via SSH shell it installs without a glitch. Same user, same folder [07:09] jbpros has joined the channel [07:10] TheJH has left the channel [07:10] TheJH has joined the channel [07:11] Elems1 has joined the channel [07:11] admc has joined the channel [07:11] der|: Ok guys, project is Up [Node-tpl] http://code.google.com/p/node-tpl/ [07:11] der|: Use, Abuse and Enjoy :) [07:12] jMerliN: does require block? [07:12] ckknight: jMerliN: yes, it's synchronous [07:12] TheJH: jMerliN, sometimes :) [07:12] mattcodes has joined the channel [07:12] luckysmack: php short tags. what if i need to use php somewhere? [07:13] ckknight: TheJH: I thought always. [07:13] Nexxy: luckysmack, what about short tags? ;o [07:13] TheJH: ckknight, well, you could say that loading from in-memory cache isn't blocking [07:13] ckknight: well, only because it's so fast. [07:13] ckknight: it's still synchronous. [07:13] luckysmack: are php short tags [07:13] ckknight: luckysmack: you should keep your php and JS in separate files. [07:13] Nexxy: luckysmack, ya but I don't understand the relevance [07:14] luckysmack: yea good point. im still new to this all. i just started yesterday. [07:14] luckysmack: im coming from php [07:14] luckysmack: so more curious than anything else [07:14] der|: luckysmack: that makes 2 of us :) [07:14] gvz has joined the channel [07:14] luckysmack: sweet [07:15] luckysmack: is this the only templating system for node so far? or are there others.? [07:15] der|: guys, check out the templating system and tell me what do you think :) http://node-tpl.googlecode.com/files/node-tpl-0.1.zip [07:15] der|: luckysmack: nope, check out https://github.com/joyent/node/wiki/modules [07:15] vortec2 has joined the channel [07:16] luckysmack: ah, a lot [07:16] ckknight: der|: " ?> [07:16] der|: ckknight: broken already ? lol [07:16] der|: that's a record heh [07:16] ckknight: I like to find bugs [07:16] CIA-65: node: 03Ryan Dahl 07master * r21964fa 10/ (3 files in 3 dirs): Upgrade libuv to b2ed24d - https://github.com/joyent/node/commit/21964fa779abd7d089cc843f67ffc1c8e8636e40 [07:16] der|: me too heh [07:17] Brandonn has joined the channel [07:17] luckysmack: haha [07:17] der|: ckknight: thanks for that one... It's late here, I just wanted to get it out just in case somebody might find it useful [07:17] luckysmack: so question, granted i havent looked at many of the others yet, what the benefit ov this one over others? [07:17] ckknight: it's PHP-like [07:17] der|: luckysmack: Not much really, I just wanted to code my own version of a templating system. [07:18] ccare has joined the channel [07:18] der|: but yes, it's php like as ckknight said [07:18] ckknight: I much prefer a saner templating system where you pass in an Object that renders [07:18] ckknight: or if you have a full page, I'm a fan of jade [07:18] cryptix has joined the channel [07:19] luckysmack: yea same here. in a template id prefer that the page only has what i have passed to it. [07:20] e6nian has joined the channel [07:20] emattias has joined the channel [07:21] [CDATA[: der| you could do that with underscore template. just var _ = require('underscore'); _.templateSettings = { evaluate: //g, interpolate: //g }; [07:21] blueadept has joined the channel [07:22] der|: [CDATA[: I see... [07:22] der|: Well, reinventing the wheel is normal for beginners I guess heh [07:22] [CDATA[: yeah…you should steal there regexen [07:22] ablomen has joined the channel [07:22] jacobolus has joined the channel [07:23] unlink has joined the channel [07:23] unlink has joined the channel [07:24] sreeix has joined the channel [07:24] [CDATA[: der| it would be cool if you actually made it support some php idioms…and even leave their https://github.com/cliftonc/calipso ?? [08:49] `3E|TV has joined the channel [08:50] mscdex: Sironfoot: you could always separate out your routes and middlewares into separate files and load them in app.js instead [08:51] robhawkes has joined the channel [08:51] Sironfoot: ok, thanks mscdex [08:51] jbpros has joined the channel [08:52] Sironfoot: Any decent tutorials on how to do that? [08:52] Swizec has joined the channel [08:53] fyskij has joined the channel [08:53] mscdex: Sironfoot: just create modules that export your middleware or route(s) [08:53] mscdex: you could have a module that exports a set of routes [08:53] jbpros has joined the channel [08:53] Nuck: Question. [08:53] Nuck: How the fuck do you guys name shit. [08:54] hkjels has joined the channel [08:54] mscdex: or you could pass in your app object to the modules and have it add the route(s) directly [08:54] Nuck: I've been trying to name my startup for HOURS solid. [08:54] mscdex: Nuck: node-startup [08:54] mscdex: call it that [08:54] mscdex: :-D [08:54] Nuck: lolno [08:54] Nuck: Not a module [08:54] aoberoi: so i just used ngen to start a module because it says it sets up some best practices. the index.js i got is a one liner to include ./lib/myproj.js. is that typical for index.js? what else might go in there? [08:54] Nuck: Sironfoot: I do that a lot. [08:55] mscdex: Nuck: no, as a startup name! [08:55] Nuck: I was just doing that today, in fact. [08:55] Nuck: I export a function, then do require()(app) [08:55] mscdex: jeesh it's almost 5 [08:55] galaxywatcher has joined the channel [08:55] Nuck: And then the function takes the app thing as an argument [08:55] mscdex: time flies when you're having fun [08:56] hecticjeff has joined the channel [08:56] skm has joined the channel [08:56] aoberoi: Nuck: so you put everythingin index.js? [08:56] Nuck: huh [08:56] aoberoi: sorry i wasnt sure if u were responding to me or not [08:56] Nuck: aoberoi: WAsn't [08:57] Nuck: Was directed to sirdancealot. [08:57] Nuck: :P [08:57] Nuck: or wait [08:57] Nuck: Fail [08:57] indutny: ryah: is there any tasks for nodejs evenangelists? ;) [08:57] Sironfoot: So you pass a reference to 'app' to your modules? [08:57] Nuck: Sironfoot: After encapsulating in a function, yes [08:58] __doc__ has joined the channel [08:59] Sironfoot: cool ok, but there's no absolute recommended way to organise everything, really up to the dev, a few diff approaches etc? [08:59] juli0w has joined the channel [09:00] japj: Sironfoot: I think the node community is still figuring out best practices for doing things, so there is no absolute recommended way to organise stuff yet [09:00] Nuck: Sironfoot: You hafta organize shit yourself, you get to do all modular shit. [09:00] Me1000 has joined the channel [09:00] Nuck: I appreciate that freedom, personally :P [09:01] Nuck: Dear god [09:01] Nuck: My team is great at naming pornsites. [09:01] Nuck: We've come up with at least two in the past hour. [09:02] Nuck: But we just cannot come up with a good name for an art site. [09:02] markwubben has joined the channel [09:02] Sironfoot: Cool ok, at least I know, I'll try different things out and see what works. I know it's still early days, but would be good if Node.js came up with recommended way to organise your project, or give directions to furst time users [09:02] mscdex: indutny: get a pilot to fly around the city with a large banner behind him saying, "Node.JS Rules!" [09:02] Sironfoot: Thanks for your advice people :-) [09:02] japj: nuck: you could also pick letters from all your names and make something with that [09:03] indutny: mscdex: heh [09:03] indutny: mscdex: that's your choice [09:03] Nuck: japj: We came up with another pornsite name that way D: [09:03] Nuck: TAGP - taggable porn. [09:03] japj: Nuck: sigh [09:04] japj: Nuck: what about naps (not a porn site) ;) [09:05] mscdex: Nuck: nsfw.js? [09:05] Sironfoot: nude.js [09:07] industrial: Howcome a JSON.parse should fail (unexpected token ILLEGAL) but it validates on jsonlint.com ? [09:07] liar has joined the channel [09:07] japj: does it say where it is failing? [09:08] industrial: nvm [09:08] industrial: yeah, my fault [09:08] seivan has joined the channel [09:09] booo has joined the channel [09:10] nail_ has joined the channel [09:11] Twist9 has joined the channel [09:11] ckknight: hey industrial [09:13] luke` has joined the channel [09:13] Swizec_ has joined the channel [09:14] nail_ has joined the channel [09:15] bmaland has joined the channel [09:17] anubhaw has joined the channel [09:17] luke` has joined the channel [09:17] Twist9 has left the channel [09:18] dsirijus has joined the channel [09:19] MrTopf has joined the channel [09:19] gozala has joined the channel [09:20] mattp_ has joined the channel [09:21] hellp has joined the channel [09:22] `3rdEden has joined the channel [09:25] FireFly has joined the channel [09:26] mendel_ has joined the channel [09:29] skohorn has joined the channel [09:30] adrianmg has joined the channel [09:32] luke` has joined the channel [09:37] jomoho has joined the channel [09:39] boehm has joined the channel [09:40] socketio\test\87 has joined the channel [09:41] andrewfff has joined the channel [09:42] V1 has joined the channel [09:43] hwinkel has joined the channel [09:44] couchquid has joined the channel [09:44] socketio\test\94 has joined the channel [09:45] jakehow has joined the channel [09:45] markwubben has joined the channel [09:45] jj0hns0n_ has joined the channel [09:47] wenbert has joined the channel [09:48] wenbert: im thinking of storing data for every action made. is this possible? what persistent storage should i use? [09:55] gg777 has joined the channel [09:55] ttpva has joined the channel [09:56] ttpva has joined the channel [09:57] _bat has joined the channel [09:57] Swizec has joined the channel [09:57] radiodario has joined the channel [09:57] _bat: oh, nice - didn't know that jsonlint has a CLI. npm install -g jsonlint :) [09:58] MrWarGames has joined the channel [09:59] MrWarGames has joined the channel [09:59] MrWarGames has joined the channel [10:01] onar has joined the channel [10:02] isaiah_ has joined the channel [10:02] ph^ has joined the channel [10:02] jerrysv has joined the channel [10:02] kimico has joined the channel [10:04] jacobolus has joined the channel [10:05] Nexxy has joined the channel [10:06] rurufufuss_ has joined the channel [10:06] rurufufuss_ has left the channel [10:06] digiwano has joined the channel [10:06] duckspeaker has joined the channel [10:07] thalll has joined the channel [10:07] mehlah has joined the channel [10:07] [[zz]] has joined the channel [10:08] bogo has joined the channel [10:09] Nuck has joined the channel [10:11] michaelhartau has joined the channel [10:13] rurufufuss has left the channel [10:20] fermion has joined the channel [10:21] jbrokc has joined the channel [10:25] hij1nx has joined the channel [10:25] shanebo has joined the channel [10:26] mattcodes has joined the channel [10:29] ditesh|cassini has joined the channel [10:29] svenlito has joined the channel [10:31] StepanKuzmin has joined the channel [10:32] TrevGamingRIg has joined the channel [10:32] saschagehlich has joined the channel [10:32] TrevGamingRIg has left the channel [10:33] TrevGamingRIg has joined the channel [10:33] TrevGamingRIg has left the channel [10:34] patrickgamer has joined the channel [10:35] patrickgamer has left the channel [10:37] mendel_ has joined the channel [10:38] tmedema has joined the channel [10:39] tmedema: Does anyone know how to pause an ongoing piping action from stream A to stream B to make it resume to stream C without going through stream B? I've asked this question on the ML, http://groups.google.com/group/nodejs/t/e7eb921cf1c50405 [10:40] RushPL has joined the channel [10:44] hwinkel has joined the channel [10:46] liar has joined the channel [10:48] bmaland has joined the channel [10:50] pen has joined the channel [10:51] adrianmg has left the channel [10:53] mhauri has joined the channel [10:55] H4ns` has joined the channel [10:56] TheJH: tmedema, you stream A->B and now want to switch to A->C? [10:56] tmedema: TheJH: yes [10:56] tmedema: while the pipe is in progress [10:56] TheJH: tmedema, when exactly do you want to switch [10:56] TheJH: ? [10:57] tmedema: TheJH: after having setup the piping from A to C, while it is going on [10:57] tmedema: somewhere inbetween that process I will get information that may or may not make me want to bypass it to stream C [11:00] TheJH: tmedema, would it be ok to close B in that case? [11:00] luke`_ has joined the channel [11:00] tmedema: TheJH: hmm, if necessary, I guess.. but it could cause some unexpected behavior [11:00] tmedema: is that really necessary? it seems unrelated [11:02] markwubben has joined the channel [11:02] TheJH: tmedema, as there doesn't seem to be a way to cleanly cancel a pipe() operation, I'd have suggested to do something like "b.emit('end')" [11:03] tmedema: TheJH: hmm, alright, so if I end stream B then the piping process will stop gracefully? [11:04] TheJH: tmedema, looks that way [11:05] milani has joined the channel [11:05] tmedema: TheJH: righto, thanks a lot [11:05] materialdesigner has joined the channel [11:05] TheJH: tmedema, you can see it yourself at https://github.com/joyent/node/blob/master/lib/stream.js [11:05] tmedema: sounds like there is a need of change in API that allows for more control over the piping process though [11:05] TheJH: tmedema, sounds like a good task for a module. want to write one? [11:07] tmedema: TheJH: actually need to work on a deadline, but if it doesn't involve C then I may work on it later [11:07] TheJH: tmedema, you won't need C for that :) [11:07] mhausenblas_ has joined the channel [11:08] jeedey: I have a variable declared in the normal way (var name = [];) and then a function with a callback just after, the callback cant access the variable? but its possible in JS.. [11:08] ttpva has joined the channel [11:08] tmedema: TheJH: basically you're suggesting an updated Stream module? [11:08] TheJH: jeedey, could you gist your code? it should work [11:09] TheJH: tmedema, something like a wrapper class for Stream or so [11:09] akshatj has joined the channel [11:09] davidbanham has joined the channel [11:09] jeedey: http://pastie.org/2308514 (err line 6, cannot modify [1] of undefined) [11:10] jeedey: probably missing something simple [11:10] MonsieurLu1 has joined the channel [11:10] SuMarDi has joined the channel [11:10] asabil has joined the channel [11:11] TheJH: jeedey, that's the "the variable changed because of the loop before my callback fired" thing [11:11] jeedey: ohh [11:11] jeedey: of course [11:12] TheJH: s/because of the loop/because the loop/ [11:12] jeedey: thank you [11:12] jeedey: knew id missed something simple [11:12] jeedey: XD [11:12] TheJH: jeedey, by the way, you declared i somewhere, right? [11:13] jeedey: ya [11:14] dsirijus has joined the channel [11:15] temp01 has joined the channel [11:18] temp02 has joined the channel [11:21] juli0w has joined the channel [11:22] mhausenblas_ has joined the channel [11:22] temp01 has joined the channel [11:23] tmedema: How does one check if a writable stream is ready for writing (the kernell buffer is empty) without actually writing? [11:25] TheJH: tmedema, you keep track of the results of "write" and the "drain" event [11:26] tmedema: TheJH: right so I have to register my own boolean and listen for the drain event? I cannot listen to write's returns since I do not call write myself (other libraries do) [11:26] unlink has joined the channel [11:27] michaelhartau has joined the channel [11:27] TheJH: tmedema, do you give the stream to other libs or do they give you the stream? [11:27] tmedema: TheJH: they give me the stream [11:27] asabil has joined the channel [11:27] tmedema: or actually, nevermind that [11:27] tmedema: I give them the stream (a writable stream) [11:29] tmedema: I guess I could wrap the write function.. seems rather ugly though. [11:29] gavin_huang has joined the channel [11:29] MonsieurLu has joined the channel [11:30] TheJH: tmedema, this is somehow ugly, but it should work: `var write = myStream.write; myStream.write = function() { return this.ready = write.apply(this, arguments); }` [11:30] fyskij has joined the channel [11:30] tmedema: TheJH: yeah, that's what I mentioned too [11:30] tmedema: will do, seems like the streaming API is a bit narrow [11:31] TheJH: tmedema, ah, ok. just make sure that you really only do it on the instance and not on the prototype. [11:31] fairwinds has joined the channel [11:32] tmedema: TheJH: by the way I do not think you need to save myStream.write in write. Inside the function the write method will not have been overwritten, yet. [11:32] svenlito has joined the channel [11:32] tmedema: myStream.write = function() { return this.ready = myStream.write.apply(this, arguments); } [11:32] TheJH: tmedema, I think that you're wrong there - when the method gets called, the property already was overwritten [11:33] TheJH: tmedema, that one's gonna give you endless recursion [11:33] tmedema: TheJH: ah of course, I wasn't thinking right.. sorry & thanks [11:33] TheJH: tmedema, no problem :) [11:33] chjj has joined the channel [11:34] mhausenblas_ has joined the channel [11:35] objectiveous has joined the channel [11:36] Metal3d has joined the channel [11:36] jj0hns0n has joined the channel [11:37] Hodgestar has joined the channel [11:38] michaelhartau has joined the channel [11:38] balupton has joined the channel [11:41] jetienne has joined the channel [11:41] enotodden has joined the channel [11:45] maru_cc_ has joined the channel [11:47] CrisO has joined the channel [11:47] eee_c has joined the channel [11:51] fyskij has joined the channel [11:51] tmedema: TheJH: when I do writeStream.end() while it is being piped to I actually seem to get 'net.js:334 throw new Error('Socket.end() called already; cannot write.');' [11:52] jbpros: any suggestion for spying on the main function of a module with jasmine? I'd like to spy on calls to the function returned by something like `require('module');` but spyOn() only works on "object/property" pairs… [11:52] thalll has joined the channel [11:53] TheJH: tmedema, hmm, I don't know enough to help you with that, I think [11:54] tmedema: alright, it's a bugger to debug these streaming issues.. [11:55] ccare has joined the channel [11:55] hwinkel has joined the channel [11:57] xippi has joined the channel [11:58] liar has joined the channel [11:59] michaelhartau has joined the channel [11:59] mgrayson has joined the channel [12:00] jbpros: another way to put it: is there a way to stub "require()" with jasmine? [12:01] djcoin has joined the channel [12:01] jbpros: it looks like require is a different function in each source file. changing it to some custom function in a spec file has no effect in the "tested" file [12:02] ryanfitz has joined the channel [12:02] monokrome has joined the channel [12:02] piscisaureus has joined the channel [12:03] nail_ has joined the channel [12:04] broofa has joined the channel [12:04] jtrudeau has joined the channel [12:05] xonev has joined the channel [12:07] mike5w3c has joined the channel [12:08] scott_gonzalez has joined the channel [12:09] felixge has joined the channel [12:09] felixge has joined the channel [12:10] milani: who used node gitteh module here? [12:11] milani: I have a benchmarking question;) [12:12] liar has joined the channel [12:14] `3E|TV has joined the channel [12:14] luke`_ has joined the channel [12:14] nail_ has joined the channel [12:15] gozala has joined the channel [12:15] nail_ has joined the channel [12:18] igl1 has joined the channel [12:19] cosmincx1 has joined the channel [12:19] svend has joined the channel [12:21] agnat has joined the channel [12:22] shinuza has joined the channel [12:24] luke` has joined the channel [12:25] bradleymeck has joined the channel [12:25] brianseeders has joined the channel [12:25] pandeiro has joined the channel [12:27] ditesh|cassini has joined the channel [12:28] cccaldas has joined the channel [12:29] topaxi has joined the channel [12:29] hvgotcodes has joined the channel [12:29] hvgotcodes: how do i execute an expensive function asynchronously [12:30] TheJH: hvgotcodes, cpu expensive? io expensive? ... [12:30] TheJH has left the channel [12:30] hvgotcodes: cpu [12:30] TheJH has joined the channel [12:30] tuhoojabotti: callbacks! [12:30] tuhoojabotti: :D [12:30] TheJH: oops... damn ctrl+w [12:30] japj: javascript or c function? [12:30] hvgotcodes: tuhoojabotti: right, but is there a node function that executes teh function, and the callback, when done? [12:31] hvgotcodes: japj: javascript, in the server [12:31] TheJH: hvgotcodes, I think that you want workers [12:31] TheJH: !npm search worker [12:31] tuhoojabotti: Yeah [12:31] hvgotcodes: so one solution is to create a child process, but this is running in the node.js code itself [12:31] jhbot: packages (short format): beanstalk_worker, worker, resque, jjw, worker-pool, spawn, gearman, q-comm, webworker, gearnode, factory-worker, cluster-isolatable, kue, fibers [12:31] hvgotcodes: TheJH: its a module, no native support? [12:32] vortec2 has left the channel [12:32] TheJH: hvgotcodes, well, you could do it yourself using the native child_process api, but it's easier using a lib [12:33] vortec2 has joined the channel [12:34] hvgotcodes: TheJH: it seems child_process only runs things on the server, outside the node server [12:34] topaxi has joined the channel [12:34] hvgotcodes: is there a child process method that fires a js function that is defined in the server? [12:34] TheJH: !npm search webworker [12:34] jhbot: package webworker: An implementation of the HTML5 Web Worker API [12:36] hvgotcodes: TheJH: is that for running background threads on the node server? [12:36] TheJH: hvgotcodes, how does "var w = new Worker('foo.js');" look to you? [12:36] hvgotcodes: TheJH: not ideal, id like to define the function in the server itself [12:37] TheJH: hvgotcodes, you can't really execute a function in the background, but you can start a worker and exchange messages with it [12:37] fumanchu182 has joined the channel [12:37] hvgotcodes: how about instead of exchaning messages, callbacks get executed on success/fail? [12:38] TheJH: hvgotcodes, I'd say that that's kind of a message [12:38] ditesh|cassini has joined the channel [12:38] hvgotcodes: true [12:38] TheJH: hvgotcodes, have a look at https://github.com/pgriess/node-webworker [12:38] nail_: there is child_process.fork that is a bit like web workers [12:39] nail_: you can trigger functions on the parent on specific messages [12:39] nail_: oh wait, the parent can send messages to the child, not sure if it works the other way [12:39] TheJH: nail_, I don't think that that function exists [12:39] hvgotcodes: TheJH: yeah i was going over that [12:39] nail_: TheJH: it's in 0.5 [12:40] nail_: http://nodejs.org/docs/v0.5.3/api/child_processes.html#child_process.fork [12:40] TheJH: nail_, awww, why doesn't "latest" point there? [12:40] cafesofie has joined the channel [12:41] japj: 0.5 is unstable [12:41] nail_: it's "latest stable" probably :) [12:41] TheJH: nail_, maybe [12:41] japj: and you're right, 0.4.8 is not the latest stable either [12:41] hvgotcodes: yeah latest stable [12:41] japj: there is an issue on github related to that aswell [12:41] hvgotcodes: plus, the fork method does not support just a function with some callbacks [12:41] hvgotcodes: so its no simpler than web worker [12:42] TheJH: !issue search latest nodejs [12:42] jhbot: found issues: 5, showing the first 3 [12:42] jhbot: Issue: #1408: http://nodejs.org/docs/latest/ points at 0.4.8, should be 0.4.10 [12:42] jhbot: Issue: #495: HTTP polling broken with keep alive and GET requests [12:42] jhbot: Issue: #871: Documentation/Website: Provide route/symlink always pointing to the latest API version [12:43] mgrayson has left the channel [12:43] hvgotcodes: thanx people, gonna go with webworker [12:44] topaxi has joined the channel [12:45] davidbanham has joined the channel [12:47] ayublin has joined the channel [12:47] AaronMT has joined the channel [12:49] cafesofie has joined the channel [12:49] jeedey: using fs.readFile, opening a normal txt file and console.log shows: http://pastie.org/2308860 [12:50] RC1140 has joined the channel [12:51] TheJH: jeedey, that's how it's supposed to be [12:51] TheJH: jeedey, do you want to get back a string or so? [12:51] fyskij has joined the channel [12:51] shanebo has joined the channel [12:51] jeedey: hmmm but i always read a file using the same method, usually gives me back the string [12:51] TheJH: jeedey, then I guess that you usually specify the encoding [12:52] jeedey: nope :/ [12:52] jeedey: infact, i even copied the code that i usually use and just changed the filename [12:52] jeedey: odd [12:52] TheJH: jeedey, well, the docs say: "If no encoding is specified, then the raw buffer is returned." [12:53] jeedey: ahh thanks [12:53] jeedey: :) [12:56] pickels has joined the channel [12:57] papandreou has joined the channel [12:57] sethetter has joined the channel [13:00] booo has joined the channel [13:00] hij1nx has joined the channel [13:00] SuMarDi has joined the channel [13:01] mertimor_ has joined the channel [13:02] shanebo: hey guys, do all the sync methods just block until they're done? [13:02] TheJH: shanebo, yes [13:03] TheJH: shanebo, they block everything [13:03] shanebo: TheJH, is this a no no? [13:04] TheJH: shanebo, depends on your usecase. if you want to load a configuration file on startup, you can do it, but you can imagine that using it in a file server or so is one of the worst things you could do [13:05] TheJH: shanebo, basically, it's ok for everything that only happens once on startup or so, but for everything else, use async methods [13:05] shanebo: TheJH so if someone hits the site while a sync method is active, they just get added to the queue right? [13:06] kriszyp has joined the channel [13:06] shanebo: okay good, I'm considering using it on startup of simple flat file json loads at start [13:06] TheJH: shanebo, yes, their request will not be handled until the sync method has finished [13:06] TheJH: shanebo, that should be fine - for example, require() is blocking, too [13:07] shanebo: yeah I figured [13:08] shanebo: is there a good article on how evented IO works for newcomers like me coming from threaded environments? [13:08] okuryu has joined the channel [13:09] TheJH: shanebo, here's an introduction by ryan: http://www.youtube.com/watch?v=jo_B4LTHi3I [13:10] shanebo: yep i've seen that one [13:10] shanebo: it would be cool to design a little chart of use cases and how node.js handles certain things [13:11] boehm has joined the channel [13:11] boaz has joined the channel [13:13] keeto has joined the channel [13:13] hdon- has joined the channel [13:14] cafesofie has joined the channel [13:15] manu_ has joined the channel [13:16] idefine has joined the channel [13:16] michaelhartau has joined the channel [13:16] papandreou: hij1nx: Hi, j1nx! Hmm... I guess there really isn't any other explanation besides V8 having different perf characteristics on Linux and OSX? [13:17] low-b has joined the channel [13:18] papandreou: hij1nx: The only reason why I dug into it was that I really wouldn't like for everyone to start using new Object :) [13:18] simenbrekken: Will node search multiple levels of node_modules folders? I'm having an issue where I have a node_modules folder which contains a module that requires another module that's 3 levels up in another node_modules folder and it can't be located. [13:19] theCole has joined the channel [13:20] TheJH: simenbrekken, your module that requires the other module should have it in its own node_modules folder, too [13:20] simenbrekken: I think I might have used the node_modules folder for application libaries [13:21] TheJH: simenbrekken, what does "npm ls" say? [13:21] simenbrekken: TheJH: I get a ton of unmet dependencies warnings [13:22] TheJH: simenbrekken, what does "npm install" without arguments say? [13:22] simenbrekken: TheJH: in what directoy should I be running this [13:23] TheJH: simenbrekken, in the root folder of your project which contains a node_modules folder [13:23] simenbrekken: TheJH: I'm on EC2 atm which has a bit of a strange setup so I have a node_modules in my home directory with stuff like mongodb etc. and then an application/node_modules with application specific stuff [13:24] TheJH: simenbrekken, which npm version are you running? [13:24] simenbrekken: 1.0.15 [13:24] TheJH: simenbrekken, run it in the same folder in which you also ran "npm ls" [13:25] jarek has joined the channel [13:25] jarek has joined the channel [13:25] jvduf has joined the channel [13:25] jarek: Hi [13:25] simenbrekken: TheJH: I get an error about packagae.json [13:25] simenbrekken: package.json* [13:25] TheJH: simenbrekken, could you gist it or so? [13:26] jarek: is it possible to prevent client from opening websocket connection unless it provides a valid password? [13:26] jspiros has joined the channel [13:26] simenbrekken: TheJH: sure, want to see the entire dir structure? [13:26] jarek: I now I can ask for pass once the websocket connection has been established, but is it really the best way to do this? [13:26] TheJH: simenbrekken, I'm mainly interested in the error message [13:26] jarek: s/now/know [13:26] mbrevoort has joined the channel [13:27] TheJH: jarek, I think that it is the best way. [13:27] heavysixer has joined the channel [13:28] Epeli has joined the channel [13:29] jbrokc has joined the channel [13:29] thomblake has joined the channel [13:29] thomblake has left the channel [13:30] simenbrekken: TheJH: https://gist.github.com/7db3d4cd9ff8e5e4bb3b [13:31] TheJH: simenbrekken, could you also gist the output of "npm ls -g"? [13:32] nuck-mobile has joined the channel [13:32] jensn_ has joined the channel [13:32] simenbrekken: TheJH: i updated the gist now [13:32] jif_ has joined the channel [13:32] colinclark has joined the channel [13:33] squeese has joined the channel [13:33] TheJH: simenbrekken, sorry, but that looks just weird to me [13:33] simenbrekken: TheJH: I think my setup is somewhat borked [13:33] willwhite has joined the channel [13:35] ceej has joined the channel [13:36] thalll has joined the channel [13:36] liar has joined the channel [13:37] shinuza has joined the channel [13:38] davidsklar has joined the channel [13:38] MarkMenard has joined the channel [13:38] ngs has joined the channel [13:40] mpavel has joined the channel [13:40] hij1nx has joined the channel [13:42] AD7six has joined the channel [13:43] blup has joined the channel [13:43] malkomalko has joined the channel [13:47] ksheurs has joined the channel [13:47] hacksparrow has joined the channel [13:47] AD has left the channel [13:47] bayousoft has joined the channel [13:48] jbrokc: simenbrekken: try `ln -s /usr/local/lib/node_modules ~/.node_modules` although i have no idea what the problem is [13:48] jbrokc: and put all of those npm packages in /usr/local/lib/node_modules [13:49] Tidwell has joined the channel [13:50] japj: http://blog.nodejs.org/2011/03/23/npm-1-0-global-vs-local-installation/ [13:50] simenbrekken: I presume it's just the ec2 install that makes everything borked [13:50] simenbrekken: I might just reinstall npm completely and see how that works [13:51] AD7six has joined the channel [13:51] brolin has joined the channel [13:52] brez_ has joined the channel [13:52] bradleymeck has joined the channel [13:53] AD7six has joined the channel [13:53] Cros79 has joined the channel [13:53] simenbrekken: is there a way to make this less painful to look at? require('./../../../lib/arcgis') I used to have it in app_dir/node_modules but I think that might be a bad idea also [13:54] piscisaureus: mraleph: here? [13:54] Poetro has joined the channel [13:54] Poetro has joined the channel [13:57] MaSch: does anyone know a managesieve implementation for nodejs? can't find one .. [13:57] AD7six has joined the channel [13:59] catb0t has joined the channel [14:00] DEVX__ has joined the channel [14:01] nibblebot has joined the channel [14:02] mpavel: simenbrekken: you could try using a var path = './../../' and prepend that to any require calls [14:02] mpavel: simenbrekken: you'll still have to define that one for each file though :-/ [14:02] simenbrekken: mpavel: but I'm right in not putting stuff in node_modules right? [14:03] mpavel: simenbrekken: have no clue, but I'd assume so [14:04] mpavel: simenbrekken: you'd want to allow yourself to easily update modules - meaning you don't want to change anything, as you'll loose it on updates [14:04] max8613 has joined the channel [14:04] thalll has joined the channel [14:07] brianc has joined the channel [14:07] eastender has joined the channel [14:08] thomblake has joined the channel [14:08] Swimming_bird has joined the channel [14:09] thomblake has left the channel [14:12] adrianmg has joined the channel [14:13] smtlaissezfaire has joined the channel [14:13] arnee has joined the channel [14:13] baudehlo has joined the channel [14:14] hellp has joined the channel [14:14] aheckmann has joined the channel [14:15] zaach has joined the channel [14:16] jerrysv has joined the channel [14:16] zaach has joined the channel [14:16] gg777 has joined the channel [14:17] theCole has joined the channel [14:19] max8613 has left the channel [14:20] eee_c1 has joined the channel [14:21] sourcode has joined the channel [14:21] jslatts has joined the channel [14:22] arnee has joined the channel [14:22] eee_c2 has joined the channel [14:22] pjacobs has joined the channel [14:23] mraleph has joined the channel [14:24] eee_c has joined the channel [14:24] geetarista has joined the channel [14:27] industrial: I have a problem with Mongoose modeling; [14:27] industrial: ModelA looks like {language: 'EN', value: 'foo'}, ModelB has a field name: [ModelASchema] [14:27] shanebo: i had to put mongoose down because the docs are weak [14:28] industrial: now I'm trying to make a new B like [14:28] kevwil has joined the channel [14:28] industrial: new B({... title: [{language: 'EN', value: 'foo'}] ...}) [14:28] industrial: so just one in the array, where it could be 50 [14:29] industrial: I keep getting the error that my format is wrong; [14:29] industrial: CastError: Cast to non_object_property_load failed for value "[object Object]" [14:29] industrial: I tried it with a B initializer instead of just {} but also no go [14:29] industrial: er, A initializer [14:30] industrial: actually thats a stupid format alltogether :| [14:30] seb_m has joined the channel [14:30] industrial: should be new B({... title: {EN: 'foo'} ...}) [14:31] hij1nx has joined the channel [14:31] ryanfitz has joined the channel [14:31] baudehlo1 has joined the channel [14:32] explodes has joined the channel [14:32] jakehow has joined the channel [14:33] explodes: Anyone used the i18n package? [14:33] luke`_ has joined the channel [14:33] mikeycgto has joined the channel [14:33] mikeycgto has joined the channel [14:33] fumanchu182 has joined the channel [14:34] slifty has joined the channel [14:34] jslatts: anyone using vows 0.5.9 successfully? [14:34] AD7six_ has joined the channel [14:34] softdrink has joined the channel [14:35] brianloveswords has joined the channel [14:36] jensn has left the channel [14:38] BillyBreen has joined the channel [14:38] jasong_at_apache has joined the channel [14:39] jasong_at_apache: using expressjs, what recommendations are there to securing a restful api? [14:39] RORgasm has joined the channel [14:40] seivan has joined the channel [14:40] losing has joined the channel [14:41] industrial: https://encrypted.google.com/search?q=how+to+secure+a+restful+API [14:41] softdrink: heh [14:41] edude03 has joined the channel [14:41] softdrink: secure in what way? [14:41] softdrink: semi related, this looks handy: http://code.google.com/p/crypto-js/ [14:42] nuck-mobile has joined the channel [14:43] jasong_at_apache: i was just looking if there was anything new other than run of the mill, it seems everytime i turn around there are more libs to help me out :-) [14:43] softdrink: so true [14:44] hebz0rl has joined the channel [14:44] Neil has joined the channel [14:45] gg777 has joined the channel [14:45] alexhq has joined the channel [14:45] mbrevoort has joined the channel [14:46] ns5 has joined the channel [14:47] alexhq: Hi everyone. I need your for review https://github.com/joyent/node/pull/1409 [14:47] dmbch has joined the channel [14:47] bradleymeck: secure it the good ole fashioned way, logins, but do it l ike a hypochondriac, check IP addresses, force use of https, use nonces and salt, and of course use public/ private keys injected over the login [14:48] sw8sw8 has joined the channel [14:48] softdrink: jasong_at_apache: let me know what you decide to do, i'd be curious to know :) [14:48] fumanchu182 has joined the channel [14:48] pekim has joined the channel [14:48] jasong_at_apache: sure [14:49] eee_c has joined the channel [14:50] bshumate has joined the channel [14:50] bshumate has joined the channel [14:52] seivan_ has joined the channel [14:53] clifton has joined the channel [14:57] yhahn has joined the channel [14:59] sub_pop has joined the channel [14:59] gazumps has joined the channel [15:00] baudehlo has joined the channel [15:01] replore_ has joined the channel [15:02] random123 has joined the channel [15:04] jakehow has joined the channel [15:06] hwinkel has joined the channel [15:06] davve has joined the channel [15:06] progme has joined the channel [15:06] sonnym has joined the channel [15:06] pita has joined the channel [15:06] DevX101 has joined the channel [15:06] boaz has joined the channel [15:07] jarek: How "unstable" is node 0.5.3? [15:08] captain_morgan has joined the channel [15:08] tuhoojabotti: jarek: I have been using 0.5.0 preview and now 0.5.0 and had no problems what so ever. [15:08] jarek: does it break the old API in any place? [15:08] boaz has joined the channel [15:09] DEVX_ has joined the channel [15:09] rfay has joined the channel [15:10] TomY has joined the channel [15:10] `3rdEden has joined the channel [15:10] EyePulp has joined the channel [15:11] isaacs has joined the channel [15:11] sniper506th has joined the channel [15:11] EyePulp: yo ho ho [15:12] japj: jarek: yes some of the old api gets broken in 0.5.x (that's why it is called the unstable series) [15:12] zeiris has joined the channel [15:12] confoocious has joined the channel [15:12] confoocious has joined the channel [15:13] EyePulp: running $ npm update npm -g --verbose and it gets to the registry and hangs -- here's the full dump: http://pastie.textmate.org/private/ctufp0em4v0sstlcpbhfa Anyone else have similar issues currently? [15:14] JumpMast3r has joined the channel [15:17] ph^ has joined the channel [15:17] Determinist has joined the channel [15:18] dayo has joined the channel [15:19] dayo: Hullo there! [15:19] chjj has joined the channel [15:19] dayo: I'd like to know the default install directory for nodejs [15:20] ph^ has joined the channel [15:20] ph^ has joined the channel [15:21] mpavel has joined the channel [15:21] tjholowaychuk has joined the channel [15:22] dayo: quit [15:22] TheJH: GAH! [15:22] slajax has joined the channel [15:22] TheJH: can't he just wait FIVE MINUTES? [15:22] jbpros has joined the channel [15:23] baudehlo has joined the channel [15:24] jtsnow has joined the channel [15:25] jetienne has joined the channel [15:25] mhausenblas has joined the channel [15:26] EyePulp: heh [15:26] jerrysv: no, and no he will decry how horrible node.js is because he couldn't find the answer he needed in 4 minutes and 59 seconds [15:27] mpavel has joined the channel [15:27] lukstr: he should know better than to ask a question on IRC so early in the day :) [15:27] wavded has joined the channel [15:27] petrjanda has joined the channel [15:28] chjj: yeah, everyone is still drunk from the wild nodejs parties we have every night [15:28] EyePulp: too right [15:28] TheJH: lukstr, we clearly aren't in the same timezone [15:28] dgathright has joined the channel [15:29] lukstr: TheJH: we probably are :) [15:29] Ineentho has joined the channel [15:30] zanes has joined the channel [15:30] balaa has joined the channel [15:31] nadirvardar has joined the channel [15:31] aconbere has joined the channel [15:31] wavded has left the channel [15:32] dguttman has joined the channel [15:32] alvaro_o has joined the channel [15:33] ph^_ has joined the channel [15:33] mertimor_ has joined the channel [15:33] topaxi has joined the channel [15:34] balaa_ has joined the channel [15:34] Swizec has joined the channel [15:35] alvaro_o has left the channel [15:35] edude03: Morning fellows :P [15:35] qbert_ has joined the channel [15:35] mscdex: yo [15:35] edude03: I think I need to learn javascript before I attempt to learn node.js, otherwise I'm only going to end up with a (further) bruised forthsad [15:35] tbranyen: yep [15:35] edude03: *forehead [15:36] mscdex: edude03: https://developer.mozilla.org/en/Core_JavaScript_1.5_Guide [15:36] djcoin has joined the channel [15:36] Boris__ has joined the channel [15:36] softdrink: http://eloquentjavascript.net/contents.html [15:36] edude03: Thanks guys :) [15:37] edude03: surprisingly my issues seem to only be on the client end, well other than the fact that I don't know how to debug node apps [15:37] uchuff has joined the channel [15:37] perezd has joined the channel [15:38] pietern has joined the channel [15:38] bnoordhuis has joined the channel [15:38] JJMalina has joined the channel [15:39] Corren has joined the channel [15:39] booo has joined the channel [15:39] pixel13 has joined the channel [15:40] pixel13: hey guys, anyone run into this one before? (trying to compile v0.5.2 on CentOS 5.5) ..Python 2.7.2 w/ bz2 module confirmed: [15:40] pixel13: ./configure: line 19: 14354 Segmentation fault "${WORKINGDIR}/tools/waf-light" --jobs=1 configure $* [15:41] TheJH: edude03, there are two methods for debugging: "node debug = 2.5.2 from centos's repo? [15:46] dguttman has joined the channel [15:46] pixel13: unfortunately, no [15:46] pixel13: 2.4.3 [15:46] mike5w3c_ has joined the channel [15:46] pixel13: bnoordhuis: sigh, ran these same steps on a separate CentOS 5.5 system with no issues last night…I'll see if I can determine how to resolve the compile dependencies for "dl", etc [15:47] chjj: installing things by hand is fun! [15:47] edude03: thanks :D TheJH [15:47] cafesofie has joined the channel [15:47] pixel13: chjj: agreed ;) …it went a whole lot smoother on debian... [15:47] Cromulent has joined the channel [15:48] chjj: debian is weird sometimes, its like "oh, you want to remove gnome, how about i remove all of xorg while im at it?" [15:48] tauren has joined the channel [15:48] chjj: other than that its good ;) [15:48] jarek: has anybody here tried packaging Node.js-based app for OSX? [15:49] M_ has joined the channel [15:49] jarek: I'm not sure whether I should spawn Node.js process from Objective-C or spawn Objective-C process from Node.js [15:49] Yuffster_work has joined the channel [15:50] pixel13: ok, trying this first: yum install libtool-ltdl-devel.x86_64 gdbm-devel.x86_64 gdbm.x86_64 readline.x86_64 readline-devel.x86_64 tkinter.x86_64 tkinter26.x86_64 [15:50] jj0hns0n has joined the channel [15:51] bnoordhuis: pixel13: why are you compiling python from source? [15:52] pixel13: I've read that CentOS 5 is heavily dependent on the current system version [15:53] pixel13: …my current repos max out @ 2.4.3 …when I compiled from source last night, I got the latest version and with no issues…must just be a few dependency issues on this system [15:53] adrianmg: anyone from Amsterdam? [15:53] mikeal has joined the channel [15:54] jmalina has joined the channel [15:54] Aikar: npm ERR! TypeError: Object 0.1 has no method 'match' [15:54] Aikar: how do i fix this? [15:54] topaxi has joined the channel [15:54] jmalina has joined the channel [15:54] isaacs: Aikar: what are you doing? [15:54] Aikar: npm update socket.io and npm install socket.io neither work [15:54] isaacs: Aikar: my guess is that there's a number where npm expects a version range string. [15:55] Aikar: my nodes a lil dated, but figured it wouldnt affect it [15:55] pixel13: bnoordhuis: OK, looks like my first attempt I was able to resolve these "necessary bits" for Python: dbm gdbm readline …. now just need to figure out the "dl" you mentioned [15:55] isaacs: hm, works for me? [15:55] isaacs: $ npm install socket.io [15:55] isaacs: socket.io@0.7.7 ./node_modules/socket.io [15:55] isaacs: ├── policyfile@0.0.3 [15:55] isaacs: ├── redis@0.6.0 [15:55] isaacs: └── socket.io-client@0.7.4 [15:55] jmalina has joined the channel [15:55] bnoordhuis: pixel13: libdl should be installed by default, it's part of the base system [15:57] Aikar: guess ill try updating node [15:57] broofa has joined the channel [15:57] slifty has joined the channel [15:57] shinuza has joined the channel [15:57] sivy has joined the channel [15:58] sivy has joined the channel [15:58] topaxi_ has joined the channel [15:58] bnoordhuis: pixel13: but maybe you need to install (extra?) glibc header files [15:59] pixel13: really? [15:59] topaxi_ has left the channel [15:59] pixel13: I can try to: yum install glibc-headers.x86_64 [15:59] iammerrick has joined the channel [16:00] bnoordhuis: pixel13: give it a spin, on debian systems you need libc6-dev so i suppose that's the centos equivalent [16:00] pixel13: cool, installing now [16:01] Aikar: isaacs: ive updated node to .4.10 and reinstalled npm after deleting ~/.npm and the npm node_modules folder, it still gives this error. [16:01] isaacs: Aikar: please gist the npm-debug.log file [16:02] bzinger has joined the channel [16:02] patrickgamer has joined the channel [16:02] sethetter has joined the channel [16:02] isaacs: odie5533: hey [16:03] isaacs: odie5533: you can run npm with sudo to install global packages. tha'ts fine. [16:03] isaacs: ircretary: inst odie5533 [16:03] ircretary: odie5533: `curl http://npmjs.org/install.sh | sh` (or, if there are permission issues, you can try: `curl http://npmjs.org/install.sh | sudo sh`) [16:03] isaacs: odie5533: then do `sudo npm install blah -g` to do anything global, or just `npm install bloo` to install locally [16:03] fattytuna has joined the channel [16:05] arnee has joined the channel [16:05] pixel13: bnoordhuis: hehe…ok… so I did that, also installed a few other devel packages, now the entire list during make has been reduced to: "bsddb185 dl imageop sunaudiodev" [16:05] pixel13: hehe..funny, I read "dl" is for 32-bit only…as well as the other three [16:06] DTrejo has joined the channel [16:06] Aikar: isaacs: sent it in PM [16:06] TheJH: isaacs, is the ircretary your bot? [16:07] jbpros has joined the channel [16:07] TooTallNate has joined the channel [16:07] isaacs: TheJH: yeah [16:07] isaacs: ircretary: hello [16:07] ircretary: isaacs: Hello :) [16:08] unlink has joined the channel [16:08] isaacs: ircretary: tell TheJH how to leave messages the next time he comes back online. [16:08] ircretary: isaacs: I'll be sure to tell thejh [16:08] TheJH: ircretary: notes [16:08] ircretary: TheJH: I don't have any notes for you. [16:08] isaacs: Aikar: weird. what's `npm config ls` print out? [16:09] isaacs: TheJH: oh, i think caps may confuse her. [16:09] isaacs: if you were "thejh" she'd have a note for you [16:09] DTrejo has joined the channel [16:10] Aikar: isaacs: updated gist [16:10] thejh: ircretary: notes [16:10] hacksparrow_ has joined the channel [16:10] pixel13: bnoordhuis: Interesting…seems "dl" was deprecated since Python 2.6: http://docs.python.org/library/dl.html [16:11] DavidElias has joined the channel [16:13] isaacs: Aikar: i'm stumped. [16:13] Aikar: :( [16:13] DTrejo has joined the channel [16:13] isaacs: Aikar: oh! what's in /home/dennis/src/trunk/marketing-node/package.json? [16:13] hacksparrow_ has joined the channel [16:13] isaacs: is there a 0.1 in there that should be "0.1" (with quotes) [16:13] isaacs: maybe in the dependencies hash? [16:13] Aikar: was nothing, then i tried an npm init and that didnt work either [16:13] isaacs: maybe, "express": 0.1 [16:13] Aikar: installing to a diff dir works [16:14] isaacs: Aikar: gist that package.json [16:14] isaacs: that's the issue, i'm almost sure of it [16:14] c4milo has joined the channel [16:14] Aikar: i deleted it and it still didnt work, it was the one generated from npm init [16:14] isaacs: hm, well, that's odd [16:14] isaacs: if installing in another dir DOES work, then there's something really strange going on [16:15] isaacs: ie, something in that specific directory [16:15] Aikar: renaming my node_modules folder also worked [16:15] samsonjs has joined the channel [16:15] ekryski has joined the channel [16:15] TheJH: Aikar, what's the output of 'find -name "package.json"'? [16:16] Aikar: and isaacs , is this really desired behavior... ? I did npm install in ~/, and when i just npm installed in this projs folder, it installed back in ~/node_modules instead of making a node_modules in current dir [16:17] caolanm has joined the channel [16:17] isaacs: Aikar: it walks up until it finds a node_modules or package.json [16:17] isaacs: Aikar: git does the same thing if you rename .git and have a .git folder further up [16:18] isaacs: Aikar: like git init, npm init *always* works in the cwd [16:18] Aikar: ok something about a module that was in my node_modules folder did it [16:18] Aikar: i deleted it and it worked [16:18] isaacs: ah, ok [16:18] isaacs: any idea which module? [16:18] devongovett has joined the channel [16:18] Aikar: one i wrote for private use in this proj [16:18] Aikar: bout to go check it [16:18] isaacs: somebody's package.json has a 0.1 where a string is expected. [16:18] Aikar: yep for Version [16:18] wankdanker has joined the channel [16:19] isaacs: there you go [16:19] pekim_ has joined the channel [16:19] Aikar: maybe add error checking for that? [16:19] isaacs: 0.1 also isn't a valid "version" field. must be a specific x.y.z [16:19] isaacs: yep, on it [16:19] isaacs: thanks :) [16:19] yumike has joined the channel [16:19] Aikar: or (val + "").match [16:19] Stythys has joined the channel [16:19] febits has joined the channel [16:20] arnee has joined the channel [16:21] Aikar: isaacs: another request, could you possibly make npm play nice with svn and not remove .svn meta data folders and if that folder is deleted, svn rm it if .svn meta is found? [16:21] gr-eg has joined the channel [16:21] isaacs: Aikar: svn metadata folders are gigantic [16:21] Aikar: we still use svn at work :( [16:21] isaacs: Aikar: i don't want them in the registry [16:22] zeade has joined the channel [16:22] isaacs: they're often 5-10 times the size of the package itself. [16:22] isaacs: .git folders aren't as bad, but they're still huge [16:22] Aikar: i meant when you update a module, it deletes entire folder and replaces [16:22] vortec2 has left the channel [16:22] Aikar: i mean so when you npm update xxx, it keeps the .svn folders behind [16:22] digitaltoad has joined the channel [16:22] isaacs: Aikar: so, you want npm to leave behind the ./node_modules/foo/lib/blah/.svn? [16:22] isaacs: no way [16:22] Aikar: so that the new files will be modified, and deleted files could be shown as ! and then a simple svn st | grep ^! | xargs svn rm [16:23] Aikar: not on repo commit [16:23] Aikar: just the update process [16:23] mundanity has joined the channel [16:23] Aikar: >>> svnst [16:23] Aikar: ~ ejs [16:23] isaacs: Aikar: the problem is that update doesn't actually go through and compare each file and whatnot. it just deletes the folder and puts the new one ther.e [16:23] topaxi has left the channel [16:23] _bat has joined the channel [16:23] isaacs: Aikar: because svn litters the entire tree with its metadata, that makes it a bit hopeless. [16:23] Aikar: yeah and what i meant was walk the folders and delete all files, leaving .svn folders intact [16:23] Aikar: i know [16:23] isaacs: ugh, no [16:23] isaacs: can't do that. [16:24] Aikar: :( [16:24] isaacs: sorry, it's just too awful, for too little benefit. i'd recommend putting your node_modules in .svnignore and then managing them separately [16:24] isaacs: or use git :) [16:24] dannycoates has joined the channel [16:24] Aikar: i wonder if git-svn would handle this [16:24] DEVX_ has joined the channel [16:25] isaacs: in my experience, git-svn is horrendous for all but the most trivial svn repos [16:25] isaacs: if you have any considerable amoutn of history, it's pretty bad. [16:25] EyePulp: isaacs: trying to do a self update of npm and it's erroring - "TypeError: Cannot call method 'filter' of undefined" ring any bells? doing npm update npm -g --verbose on npm@1.0.9-1 [16:25] Aikar: well this is pretty trivial repo :P [16:25] ttpva has joined the channel [16:25] isaacs: ircretary: inst EyePulp [16:25] ircretary: eyepulp: `curl http://npmjs.org/install.sh | sh` (or, if there are permission issues, you can try: `curl http://npmjs.org/install.sh | sudo sh`) [16:25] EyePulp: I can't self-update anymore? [16:25] isaacs: EyePulp: it can as of 1.0.10ish [16:26] isaacs: maybe 11, don't remember. [16:26] EyePulp: ah [16:26] isaacs: EyePulp: when in doubt, curl|sh [16:26] _bat: Aikar I used git-svn for a year on a repo w/ a few thousand files. No svn branches, though. (I did have local git branches.) [16:27] dshaw_ has joined the channel [16:27] Cleer has joined the channel [16:27] robertfw has joined the channel [16:27] EyePulp: isaacs: danke - all better now [16:28] japj: isaacs: should npm do anything yet on node 0.5.3 on windows? (I'm not getting any output, not even --help) [16:28] hacksparrow has joined the channel [16:28] TheJH: japj, npm doesn't work on windows yet [16:29] MrTopf_ has joined the channel [16:29] ts__ has joined the channel [16:29] japj: TheJH: yeah I know, but child_process was added in node 0.5.3, so I wanted to try again [16:29] patrickgamer has left the channel [16:29] cccaldas has joined the channel [16:30] TooTallNate has joined the channel [16:30] japj: unfortunately, node debug on windows also doesnt work yet, so I cant figure out what's happening [16:30] TheJH: japj, oh, child_process works now? that's interesting [16:30] liquidproof has joined the channel [16:31] DEVX_ has joined the channel [16:31] japj: TheJH: https://github.com/joyent/node/wiki/ChangeLog 0.5.3: Windows: improvements, child_process support. [16:31] jbpros has joined the channel [16:33] brianseeders has joined the channel [16:33] joemccann has joined the channel [16:33] Determinist has joined the channel [16:34] JJMalina has joined the channel [16:34] ericmuyser has joined the channel [16:34] glimpsenirvana has joined the channel [16:35] harthur has joined the channel [16:36] sbisbee has joined the channel [16:37] zzo has joined the channel [16:38] booo has joined the channel [16:39] CIA-65: node: 03Bert Belder 07master * r91ecfbc 10/ (9 files): Remove platform_win32_winsock - https://github.com/joyent/node/commit/91ecfbcae86cabd0b6872d6cda1d7a640b0d5737 [16:39] CIA-65: node: 03Bert Belder 07master * ra962cca 10/ (deps/v8/src/platform-cygwin.cc deps/v8/src/platform-posix.cc): [16:39] CIA-65: node: V8: Don't use mprotect on Cygwin as virtual memory is managed directly via WinAPI calls. [16:39] CIA-65: node: Upstream fix: http://codereview.chromium.org/7549009 - https://github.com/joyent/node/commit/a962cca390fed391ccf9bd0eb72617524470a970 [16:39] markdaws has joined the channel [16:40] aoberoi has joined the channel [16:40] isaacs: japj: expect npm-on-windows some time around christmas. [16:41] isaacs: it will be months, not weeks. [16:41] isaacs: sorry [16:41] japj: ooo a christmas present [16:41] vipaca has joined the channel [16:41] vipaca has joined the channel [16:41] fyskij has joined the channel [16:42] ctide: isaacs: is there some way to programmatically check whether all the npm packages for a given app are currently updated / installed? [16:42] japj: isaacs: if there is a plan/todo list, maybe post on the npm mailinglist for people that are willing to help? [16:42] isaacs: japj: already did :) [16:42] ctide: we have an open source project that a lot of non-node people interface with and consistently forget to npm install after pulling down from git [16:42] ctide: trying to alleviate the 'random' error messages they get [16:42] isaacs: ctide: i think there's some way to do this wiht npm.commands.ls [16:42] perezd has joined the channel [16:42] ekryski has joined the channel [16:43] ctide: k, i'll poke around there [16:43] ctide: thanks [16:43] isaacs: ctide: but a cli "npm verify" or something would be nice. post an issue if you'd like. [16:44] cccaldas has joined the channel [16:45] japj: isaacs: when did you post it? I think I only saw threads about people complaining that it wasnt working yet, or I just missed it somehow [16:46] thurmda has joined the channel [16:49] thurmda: Anyone use node-canvas? I'm having a problem installing a dependency for it. [16:49] thurmda: sudo apt-get install libjpeg8-dev => E: Package libjpeg8-dev has no installation candidate [16:49] adrianmg has joined the channel [16:50] thurmda: I'm on Ubuntu 10.10 [16:50] matomesc has joined the channel [16:51] thurmda: assuming I need to update repos (/etc/apt/sources.list) but not sure what to add [16:51] bartt has joined the channel [16:52] meandi has joined the channel [16:53] svend has joined the channel [16:53] gozala has joined the channel [16:54] captain_morgan has joined the channel [16:54] jmoyers has joined the channel [16:55] liquidproof has joined the channel [16:55] sandropadin has joined the channel [16:55] Me1000 has joined the channel [16:56] ambroff has joined the channel [16:57] jbrokc has joined the channel [16:58] ptlo has joined the channel [16:58] theCode has joined the channel [16:58] jakehow has joined the channel [16:59] kjeldahl_ has joined the channel [16:59] fangel has joined the channel [16:59] te-brian has joined the channel [17:00] erictj has joined the channel [17:00] mehlah has joined the channel [17:00] greg has joined the channel [17:02] captain_morgan has joined the channel [17:03] TheJH: is there a good, simple way for creating GUIs outside of the browser with node yet? [17:03] ekryski1 has joined the channel [17:03] heavysixer has joined the channel [17:04] AvianFlu has joined the channel [17:05] newy has joined the channel [17:06] TheJH: hmm, `chromium-browser --app=""` looks good, no address bar or anything [17:06] mikeal has joined the channel [17:06] soolwan has joined the channel [17:07] mikeal has joined the channel [17:07] Aikar: grr. people shouldnt be putting node_modules on .gitignore :/ [17:07] Aikar: let npm handle ignoring that >< [17:07] ale has joined the channel [17:07] TheJH: Aikar, why not? npm can't force git to ignore it, right? [17:07] ale: Hello, how can I create private chat server? Like this IRC, with private rooms, by accessing the url? Let's say my.page/hK2ako is a room, how can I make the messages from that room, to stay on that room? [17:08] thalll has joined the channel [17:08] bentkus: ale: you read up how to do that and create it [17:08] bentkus: whats the problem [17:09] Vertice has joined the channel [17:09] Aikar: TheJH: well in my case I wanted to commit to MY git repo but i just realized that the deps didnt get commited. I want my deps in my VCS, i dont want to have to npm install anywhere i check out. [17:09] Aikar: everywhere* [17:09] ale: bentkus: I don't know how to do it [17:09] Aikar: or at very least, the .gitignore should be in the .npmignore file :P [17:10] TheJH: Aikar, I think that committing node_modules is pretty bad style [17:10] negrete has joined the channel [17:10] Aikar: well, I strongly disagree with you [17:11] Aikar: requiring to use npm to install when deploying to production is bad [17:11] AvianFlu: why? [17:11] Aikar: coe should be a bundled "package" [17:11] Aikar: npm wont be installed on my prod servers [17:11] Aikar: why should it be? [17:11] TheJH: Aikar, when your dependencies get updated, you'll have to make a "deps update commit" or something like that [17:11] Aikar: and? [17:11] isaacs: Aikar: i agree with you [17:11] Aikar: thats extremely common [17:12] isaacs: add your deps to bundleDependencies, and then do `npm package` and you've got a tarball. [17:12] isaacs: Aikar: shasum it, put it on a dist server, whatever. [17:12] Aikar: i use npm as a way to get and update modules in order to commit to my projects repo. [17:13] Aikar: TheJH: and for the reason i just encountered mainly. I commited updated modules, then checked out on another location, and i was missing a ton of dependencies [17:13] TheJH: Aikar, you just do "npm install" without args and everything is there. what's the problem? [17:14] qbert_ has joined the channel [17:14] ttpva has joined the channel [17:15] perezd: is there a proper way to do gzip compression with non-static assets + expressjs ? [17:15] Aikar: TheJH: why? for what gain? [17:15] Aikar: I think that "style" is silly [17:15] TheJH: Aikar, besides, npm packages with native code may have to be recompiled anyway if you put your stuff on another pc [17:16] tjholowaychuk: perezd gzip options for node are not too great right now [17:16] tiglionabbit has joined the channel [17:16] Aikar: thats fine, id expect to recompile binary, which is something youd make in your projects planning [17:16] mbrevoort has joined the channel [17:16] Aikar: but like this case socket.io [17:16] perezd: tjholowaychuk: this makes me a sad panda [17:16] tjholowaychuk: me too [17:16] Aikar: its pure js [17:16] tjholowaychuk: sad pandaaaa [17:17] perezd: whats wrong with gzip and node? [17:17] tjholowaychuk: perezd we had some stuff in connect a while back but they were really messed up. I haven't looked into it much the last few months though [17:17] blup has joined the channel [17:18] newy has joined the channel [17:18] Aikar: TheJH: also it gives people more options for using your libs. If you checkout my modules off my github, they have all their deps already included, so you can install with npm or a git submodule. [17:18] ryanj has joined the channel [17:18] isaacs: Aikar: using subversion makes this very difficult [17:18] aoberoi has joined the channel [17:18] geetarista has joined the channel [17:19] Charuru has joined the channel [17:19] isaacs: Aikar: since it requires carefully folding folders in on all changes. [17:19] isaacs: Aikar: most of what you're describing is pretty easy in git [17:19] colinclark has joined the channel [17:19] juli0w has joined the channel [17:19] brainproxy: isaacs: what is the reason for dropping AMD compatibility in node's modules API? because the "compatibility" was limited and so might lead to confusion? [17:20] isaacs: brainproxy: see the very long mailing list thread on nodejs-dev [17:20] brainproxy: ah okay [17:20] darshanshankar has joined the channel [17:20] isaacs: brainproxy: there are userland moduels for it [17:20] CrisO has joined the channel [17:21] visudo has joined the channel [17:21] mbrevoort: isaacs: hey, is there a target timeline for npm to support windows, post node v0.6? [17:22] isaacs: mbrevoort: christmas [17:22] EvRide has joined the channel [17:22] isaacs: mbrevoort: hey, i wanted to ask you, can you replicate from http://admin.npmjs.org:5984/registry? [17:22] mbrevoort: isaacs: a Christmas miracle! [17:22] mpavel has left the channel [17:23] mbrevoort: isaacs: let me try [17:23] Aikar: isaacs: works fine for me on svn, the file structure you want in prod is pretty much what is in your VCS system. I use npm solely as a way of installing and updating modules. its non existent when it comes to deploying.. update modules, commit changes (which was my problem i used git-svn to solve since .svn dirs getting deleted blows stuff up) [17:23] isaacs: i see [17:23] isaacs: sure, that sounds like an ok workaround [17:24] isaacs: Aikar: another solution is to *not* check your deps into your vcs, but install them locally, and list them in your `bundleDependencies`. Then, for deployment, on your dev box do `npm pack`. This will create a tarball containing your code and all your deps. put that in places. [17:24] hebz0rl has joined the channel [17:24] isaacs: with that, you can then get a shasum, do a git/svn tag, etc. [17:25] zeiris has joined the channel [17:25] isaacs: you can even take that tarball and install it with npm on your target machine, but in such a way that there's no registry whatevers, since everything is bundled. [17:25] isaacs: or just rsync and unpack it. [17:26] kjeldahl_ has joined the channel [17:28] Aikar: isaacs: i really dont see the gain, id rather have the source in a pristine state you know works. what if NPM registry is down when you go to deploy? or what if internet access is down at work but internal network stuff still works. that kind of stuff would block you. when you could just host it in your own VCS and it should 'just work' [17:28] AvianFlu has joined the channel [17:28] soolwan has joined the channel [17:28] mikeal has joined the channel [17:29] jhurliman has joined the channel [17:29] isaacs: Aikar: that's the whole point, the deployment wouldn't depend on npm's registry, because all deps are bundled. [17:30] isaacs: Aikar: but the best kind of deployment system is the one that's working, so don't touch it if it is :) [17:30] rfay has joined the channel [17:30] AvianFlu: btw isaacs, you should totally have a bunch of "Fork or GTFO" bumper stickers made [17:30] zmbmartin has joined the channel [17:31] japj: "it ain't broken if you didn't fork it"? [17:31] poptartinc has joined the channel [17:32] poptartinc: Does anybody have any magical tips for debugging a https.method call that just fails without saying anything [17:32] reid has joined the channel [17:33] mjr_ has joined the channel [17:34] pixel13: python tools/waf-light --jobs=1 configure = Segmentation fault :( ...sigh. [17:34] pixel13: Tried Python 2.5.6, 2.7.2 … same result. I _must_ be missing something obvious here... [17:34] edude03: With socket.io, will connect() use the same host and port as the script was loaded from automatically? [17:34] AvianFlu: the first thing that comes to mind is "try 2.6", pixel13 [17:35] pixel13: …two CentOS 5.5 boxes, one worked perfectly, the other did not [17:35] pixel13: AvianFlu: why not :) I'll try it now…the other CentOS 5.5 box worked great w/ 2.7.2 [17:35] mikeal has joined the channel [17:35] pixel13: …my hunch is some unresolved Python dependency [17:35] AvianFlu: I think there's a node issue somewhere about something like this [17:35] AvianFlu: I seem to remember having seen it recently [17:35] japj: pixel13: are there no official python 2.5+ rpms for centos? [17:36] pixel13: unfortunately, no …I found _one_, but it was 2.5.1 [17:36] pixel13: again, 2.7.2 from source worked just fine (once I had bz2-devel present) on the other machine... [17:36] brianloveswords has joined the channel [17:36] dgathright has joined the channel [17:37] japj: pixel13: did you try the epel repository? [17:37] japj: http://download.fedora.redhat.com/pub/epel/5/i386/repoview/python26.html [17:38] skohorn has joined the channel [17:38] pixel13: japj: hmm.. no, I don't have that one configured actually [17:38] japj: http://fedoraproject.org/wiki/EPEL [17:38] pixel13: ooh nice: http://download.fedora.redhat.com/pub/epel/5/x86_64/repoview/python26.html [17:38] pixel13: x86_64 [17:39] pixel13: I've read that yum uses python itself, and that upgrading python _via_ yum in CentOS 5 can actually break a few things… [17:39] yumike has joined the channel [17:40] pixel13: so long as the .rpm doesn't overwrite the system default python, this may also work [17:40] japj: I think python26 is a side by side install (not replacing the python that is the default in centos) [17:40] pixel13: ok..I'll test in dev first… meanwhile, from source, I keep hitting: Failed to find the necessary bits to build these modules: dl sddb185 imageop sunaudiodev [17:41] pixel13: it's the "dl" I haven't been able to resolve, but that may be since I'm x86_64 [17:42] figital has joined the channel [17:43] admc has joined the channel [17:43] pixel13: AvianFlu: no go with 2.6.7 from source = ./configure: line 19: 31324 Segmentation fault "${WORKINGDIR}/tools/waf-light" --jobs=1 configure $* [17:44] AvianFlu: yeah, that was a guess >.< [17:44] pixel13: worth the shot [17:45] mbrevoort has joined the channel [17:45] AvianFlu: and you're certain you had it using the right python? [17:45] Xano has joined the channel [17:45] materialdesigner has joined the channel [17:45] pixel13: yep [17:45] pixel13: python -V = 2.6.7 [17:45] AvianFlu: I've made the mistake a few times in my life of installing a second interpreter version and then still running the first >.< [17:46] aheckmann has joined the channel [17:46] pixel13: japj: tried the 2.6.5 rpm, and got this: error: Failed dependencies: libffi.so.5()(64bit) is needed by python26-2.6.5-6.el5.x86_64 [17:46] mikeal has joined the channel [17:46] japj: did you enable the epel repository and did a yum install python26? [17:46] pixel13: no, downloaded the .rpm and did rpm -Uhb [17:47] pixel13: -Uhv * [17:47] pixel13: i'll try the yum route now to resolve dependencies... [17:49] Aikar: how do you broadcast in socket.io from the server and not a client connection? [17:49] patcito has joined the channel [17:49] ptlo has joined the channel [17:49] stepheneb has joined the channel [17:49] pietern_ has joined the channel [17:49] Aikar: socket.emit() ? [17:49] Aikar: dont seem to be getting the msg :( [17:51] Aikar: ok so im not connecting. [17:51] pixel13: japj: ok, the yum route worked via EPEL... [17:51] pixel13: unfortunately, python26 tools/waf-light --jobs=1 configure = Segmentation fault [17:51] losing has joined the channel [17:51] pixel13: …even though "import dl" is successful via this build... [17:51] pixel13: so weird [17:53] japj: pixel13: you have 2 centos machines? (are they identical?) [17:53] pixel13: I have a few…only two are identical, unfortunately not the two in question [17:54] pixel13: all are CentOS 5.5 / x86_64 [17:54] japj: pixel13: you could try building an rpm on a machine that is working. I think there are nodejs spec files available if you google them [17:55] Determinist: guys, is there any kind of exception you know that won't get captured by process.uncaughtException? [17:55] Determinist: any kind of circumstance where that's possible? [17:55] pixel13: japj: on the first machine, it just worked…it was so simple. Python 2.7.2 from source after fetching bz2-devel, and node v0.5.1+ has always built great [17:55] japj: it could be that the machine has a physical problem (maybe test the memory?) [17:55] bradleymeck has joined the channel [17:55] hlindset has joined the channel [17:56] Determinist: pixel13: using Lion? :P [17:56] bartt has joined the channel [17:57] dgathright has joined the channel [17:57] sweetd has joined the channel [17:57] ttpva has joined the channel [17:57] jbpros has joined the channel [17:57] dreamdust has joined the channel [17:58] igl1: what can i do against the persistent boner node is giving me? its really getting in the way [17:58] pixel13: Determinist: yes! but not for this node build ;) [17:58] dreamdust: lol [17:58] pixel13: japj, interesting thought… hmmm [17:59] Determinist: pixel13: reason i'm asking is because of that automatic text replacement. it replaces ... with the unicode char :P [17:59] mikeal has joined the channel [17:59] altamic has joined the channel [17:59] mbrevoort: isaacs: so from my office, port 5984 is being blocked for some reason. out network security team sometimes takes an over aggressive approach. [17:59] juli0w has joined the channel [18:00] japj: pixel13: there is a thing called memtest86 [18:00] nnnnathann1 has joined the channel [18:00] pixel13: Determinist: oh cool, didn't know that. [18:00] pixel13: japj: safe to run on an active, Production box [18:00] Determinist: pixel13: not so cool if you're writing code and it auto replaces something for you :) [18:00] Determinist: pixel13: good luck debugging that shit :D [18:00] japj: pixel13: ah no, it is a bootable thing [18:00] japj: pixel13: sry ;( [18:00] pixel13: oh ok [18:01] cccaldas has joined the channel [18:01] Tobsn has joined the channel [18:01] nnnnathann1: hello. I'm pretty new to lower-level type networking things. I am curious as to whether or not you guys think it would be possible to make serving a sequence of static images more efficient using node explicitly for the purpose of reducing the number of requests necessary... [18:02] jobim: I have a beginner question if someone has a moment. I'm reading the example at the top of page 21 in Hands-On Node.js at http://nodetuts.com/pdf/handson-nodejs-sample.pdf [18:02] Determinist: nnnnathann1: what's wrong with spriting? is the client HTML? [18:03] TheJH: nnnnathann1, for just serving static stuff, node isn't the fastest thing that exists, have a look at gatling and nginx [18:03] jobim: On line 15, the timeout parameter is passed to the start function, but in the declaration it takes no such parameter. How is that possible? [18:04] nnnnathann1: Determinist: spriting is not possible due to the nature of the images, they are (relatively) high-res and number in the hundreds (an image sequence) [18:04] nnnnathann1: TheJH: thanks, I will take a look at those [18:04] TheJH: jobim, looks like a mistake [18:04] jobim: That's what I thought too. [18:04] Tobsn: mornin [18:04] jobim: I'm new to JS but couldn't figure out a way that was possible. [18:04] kriszyp has joined the channel [18:05] TheJH: nnnnathann1, gatling is crazy fast and small while nginx is (afaik) a little bit slower but has more features [18:05] Determinist: nnnnathann1: and how did you think of doing that then? i mean, if you're going to load up hunders of images, your problem is not the format of how you're transferring them but the kind of load you're putting on the client DOM. again, assuming we're talking about a browser client. [18:05] erictj has left the channel [18:05] hij1nx has joined the channel [18:06] nnnnathann1: Determinist, indeed, but I don't actually load them into the DOM (display tree). The efficiency is very nice once they have loaded, but loading them images request after request makes it very slow at the start. [18:06] Determinist: nnnnathann1: sorry, but again, are we talking about a browser here? display tree sounds flashy :P [18:06] nnnnathann1: Determinist, I was thinking that maybe I could use node to combine the files data server side and transmit it in a single request, then parse on the other side? [18:07] nnnnathann1: Determinist, no I just meant I don't display the images, it's all in JS [18:07] nnnnathann1: TheJH, I haven't even heard of gatling, that sounds interesting [18:07] Determinist: nnnnathann1: ah. well, if you don't mind letting go of old browsers, you could use base64 to combine images into a long ass string and then reparse on the browser. [18:07] nnnnathann1: Determinist, yeah I took a look at that but I have to support back to IE7 :\ [18:08] Determinist: nnnnathann1: i feel ya brother. [18:08] Determinist: nnnnathann1: keep alive connection? [18:08] Determinist: not gonna help tho. [18:08] nnnnathann1: Determinist, being a rookie, I got excited about that type of thing... that's mainly why I was asking about it [18:09] nnnnathann1: Determinist, sounds like I would have to send the requests anyway though? [18:09] Determinist: yeah, exciting and all, but you still have a problem with support on the browser. IE7 being the problem, mostly. [18:09] dnjaramba has joined the channel [18:09] sw8sw8 has left the channel [18:09] levity_island has joined the channel [18:09] Determinist: flash can do what you're after pretty easily [18:09] nnnnathann1: Determinist, ahh, didn't realize that was only for modern browsers [18:09] nnnnathann1: Determinist, yeah maybe I will have to break down and flash it up :\ boo [18:10] Determinist: not everything, dude, just the part where you're loading up the images. i'm willing to bet you're building a gallery of sort [18:10] DEVX_ has joined the channel [18:10] jacobolus has joined the channel [18:11] nnnnathann1: Determinist, interesting idea! no not a gallery per se, I do 3D renderings and need a way to display rotatable turntables [18:12] mikeal has joined the channel [18:12] robertfw has joined the channel [18:12] asabil has joined the channel [18:12] AvianFlu_ has joined the channel [18:13] nnnnathann1: Determinist, thanks for the input man! You too TheJH, catch yo uguys later, I will be afk for a bit [18:14] brez_ has joined the channel [18:16] ericmuyser has joined the channel [18:18] Determinist: np :P [18:19] zaach has joined the channel [18:20] tuhoojabotti: Huge wall of text. :D [18:21] perezd: weird, every other request to my expressjs server throws: Error: Can't use mutable header APIs after sent. [18:22] harth has joined the channel [18:22] perezd: I have body and cookie parser enabled as middleware [18:22] loob2 has joined the channel [18:23] mikeal has joined the channel [18:23] bnoordhuis: perezd: you're trying to read/write http response headers after res.writeHead()'s been called [18:23] bnoordhuis: perezd: it's a pending bug [18:24] tommyvyo has joined the channel [18:24] feychenie has joined the channel [18:25] tjholowaychuk: perezd can i see the config? [18:25] tjholowaychuk: might be an ordering issue [18:25] tilgovi has joined the channel [18:25] tilgovi has joined the channel [18:25] perezd: tjholowaychuk: I think I call my callbacks too many times... [18:25] perezd: I think thats a bug on my end [18:25] tjholowaychuk: ah [18:26] kartmetal_ has joined the channel [18:26] feychenie: hi there, anyone here that could help with a node + mongodb + replicasets auto-failover issue ? [18:26] Fuu`` has joined the channel [18:27] perezd: is there any particular state sharing between module requires I need to be aware of "resetting" ? [18:27] wavded has joined the channel [18:27] isaacs has joined the channel [18:27] wavded has left the channel [18:28] stisti has joined the channel [18:28] ippa has joined the channel [18:28] necrodearia has joined the channel [18:30] dnjaramba has joined the channel [18:31] maushu has joined the channel [18:32] feychenie: ok so let's expose my issue, i'll see if someone could help. I've a working node app that on start connects to a mongodb replicaset, and successfully detects the right master. When i kill the master mongodb instance, one of the slaves becomes the new master. But node still tries to connect to the old master, and doesn't choose the new one. [18:32] feychenie: And all my writes fail [18:32] samsonjs has joined the channel [18:33] scott_gonzalez has joined the channel [18:33] wink_: feychenie: thats a problem with your mongo driver [18:33] isaacs: AvianFlu: that's a great idea. [18:33] aconbere has joined the channel [18:33] isaacs: AvianFlu: it's a good bumper-sticker sized thing, too [18:33] AvianFlu: definitely [18:33] AvianFlu: and after how many times I saw it retweeted yesterday, I think it would catch on [18:34] hlindset has joined the channel [18:35] harthur has joined the channel [18:35] feychenie: wink_: yes i guessed that. It's supposed to be a closed issue on the drivers repo, and i've got the latest version. So i was wondering if someone here is working with node + mongo+replicasets to see if we could compare our code [18:35] cccaldas has joined the channel [18:35] balaa has joined the channel [18:36] seb_m has joined the channel [18:40] aconbere has joined the channel [18:40] theCode has joined the channel [18:41] d0na1d has joined the channel [18:41] tristanseifert has joined the channel [18:43] DEVX_ has joined the channel [18:44] samsonjs has joined the channel [18:45] arnee has joined the channel [18:49] dherman has joined the channel [18:49] devongovett has joined the channel [18:50] thalll has joined the channel [18:52] mikeal has joined the channel [18:53] juli0w has joined the channel [18:54] DEVX__ has joined the channel [18:55] cconstantine: how reasonable would it be for 100k clients to be connected to a single node server with very little traffic? [18:55] cconstantine: or 200k [18:56] skm has joined the channel [18:57] thurmda: tjholowaychuk: I'm trying to install prereqs for node-canvas and getting an err. On ubuntu 10.10 >apt-get install libjpeg8-dev => E: Package libjpeg8-dev has no installation candidate [18:57] thurmda: I'm assuming I need to update /etc/apt/source.list to add a repo... but which? [18:57] thurmda: any advice? [18:57] tjholowaychuk: thurmda i just install from source so i dont have any package manager advice really [18:57] booo has joined the channel [18:57] thurmda: k [18:57] cconstantine: thurmda, and I'm stuck on 10.4 :( [18:58] thurmda: https://github.com/LearnBoost/node-canvas/wiki/Installation---Ubuntu [18:58] thurmda: :( [18:58] thurmda: any clue where the source is for that lib? [18:58] pixel13: thurmda: I was able to snag it just fine on Ubu 10.10 [18:59] pixel13: Get:2 http://us.archive.ubuntu.com/ubuntu/ maverick/main libjpeg8-dev amd64 [18:59] thurmda: :) [18:59] bradleymeck: cconstantine unreasonable for a single server, breaking 32k connections is bad juju on most tcp ip stacks [18:59] pixel13: are you on amd64 ? [18:59] DEVX_ has joined the channel [18:59] tristanseifert has joined the channel [18:59] cconstantine: bradleymeck, thanks. I have an idea, but we're already at some scale :) [19:00] cconstantine: bradleymeck, so, keep it around... 15k connections/server? [19:00] madsleejensen has joined the channel [19:00] temp01 has joined the channel [19:00] elifou has joined the channel [19:00] bradleymeck: just spread it across multiple machines, but yes 15k is ideal 2byte port address thing is pretty much standard for tcp/ip which can go up to 32k but ... problems when you get up there [19:01] sethetter has joined the channel [19:01] thurmda: i686 [19:01] cconstantine: bradleymeck, sweet, thanks [19:01] elifou has joined the channel [19:02] hlindset has joined the channel [19:03] _bat has joined the channel [19:03] pixel13: when I run: apt-cache show libjpeg8-dev it states: Architecture: amd64 [19:03] pixel13: thurmda: could be an issue per your processor... [19:04] Ramosa has joined the channel [19:04] thurmda: trying another machine... [19:04] walkah has joined the channel [19:04] walkah has joined the channel [19:06] BillyBreen has joined the channel [19:06] andihit has left the channel [19:06] greg has joined the channel [19:06] pixel13: thurmda: what about http://archive.ubuntu.com/ubuntu/pool/main/libj/libjpeg8/libjpeg8-dev_8b-1_i386.deb ? [19:06] kartmetal_: cconstantine, the 32k connection "limit" is per IP address. You could serve multiple IP# from one physical server. [19:07] mbrevoort has joined the channel [19:07] elifou: i have a dependency in my node app that keeps throwing an error on my machine but not when i install it on a clean vm, leading me to try to figure out what's wrong on my machine. the module was installed locally so it's easy to remove and reinstall. i don't think that's the problem. i used nvm to install node (v0.4.9). i had other versions installed but then i tried nuking it and only installing that version. [19:08] elifou: is there any secret files that anyone has found sneaking up on them using nvm? [19:08] elifou: the particular problem is that i'm using jsdom and getting an htmlparser error (a dependency) [19:09] thurmda: interesting : Selecting previously deselected package libjpeg8-dev. dpkg: considering removing libjpeg62-dev in favour of libjpeg8-dev ... dpkg: no, cannot proceed with removal of libjpeg62-dev [19:09] balaa has joined the channel [19:10] elifou: alternatively, has anyone encountered this jsdom error using it in node and solved the problem: WARNING: No HTML parser could be found. \n Element.innerHTML setter support has been disabled \n Element.innerHTML getter support will still function \n Download: http://github.com/tautologistics/node-htmlparser [19:10] crodas has joined the channel [19:11] digitaltoad has joined the channel [19:12] pixel13: can you manually remove libjpeg62-dev first, via apt-get? [19:12] pixel13: or is that what you just tried? [19:12] CIA-65: node: 03Mikeal Rogers 07master * r48dcb90 10/ (9 files in 2 dirs): Update tests for http2. - https://github.com/joyent/node/commit/48dcb905f6cfc44a8bbbddfa3d1c25415b2b5287 [19:12] puffpio has joined the channel [19:12] whitman has joined the channel [19:13] cconstantine: kartmetal_, interesting. this would be on ec2, so it might just be easier to spin up a ton of micro instances [19:13] devongovett has joined the channel [19:14] sleeplessinc has joined the channel [19:14] hellp has joined the channel [19:15] odie5533: How long have the Windows binaries of node been available? [19:16] djcoin has joined the channel [19:16] lukstr: odie5533: couple weeks [19:17] thurmda: fuck yeah! [19:17] thurmda: that was it [19:17] thurmda: thanks pixel13 ! [19:17] yozgrahame has joined the channel [19:18] hwinkel has joined the channel [19:20] Draggor: So my npm won't find anything when searching [19:21] pixel13: thurmda: sweet ;) [19:21] pixel13 has left the channel [19:21] Kingdutch has joined the channel [19:21] isaacs: Draggor: what version of npm? [19:21] towski has joined the channel [19:21] isaacs: Draggor: and how are you searching? [19:22] fangel has joined the channel [19:23] Draggor: npm search whatever [19:23] Draggor: version is... 0.2.19 [19:24] markbao has joined the channel [19:25] Draggor: Also, was there an OpenCL project on node by any chance? [19:25] juli0w has joined the channel [19:25] Aiden has joined the channel [19:25] odie5533: Does the Windows nodejs have any limitations? [19:25] rfay has joined the channel [19:25] odie5533: Or is all Node code now portable? [19:25] bradleymeck: still limits [19:26] odie5533: What limits? [19:26] bradleymeck: various, child processes, pipes, some more specific things like fs.watch, os.* [19:26] odie5533: Is there a list somewhere of what does/doesn't work? [19:27] pixel13 has joined the channel [19:27] odie5533: Are popular modules like socket.io compatible? [19:27] bradleymeck: odie5533 it changes pretty often, and there are plenty of issues. Some are, but if they use C++ they are not [19:28] heavysixer has joined the channel [19:28] bradleymeck: easier to try on a recent release, but it wont be production ready until 0.6.x so id wait [19:29] mehlah has joined the channel [19:29] Aikar: hmm, i cant kill logging from socket.io [19:29] odie5533: Wait to start using the windows port you mean? [19:29] Aikar: var io = socketio.listen(app); [19:29] `3rdEden: Aikar read the wiki [19:29] Aikar: io.set('logger', false); [19:29] Aikar: i am [19:29] Aikar: io.set('log level', 0); [19:30] Aikar: even with both of those it still logs [19:30] `3rdEden: yes that one [19:30] `3rdEden: io.set logger false will break your app [19:30] bradleymeck: odie5533 if you want to use the mingw port yea [19:30] `3rdEden: https://github.com/LearnBoost/Socket.IO/wiki/Configuring-Socket.IO [19:30] `3rdEden: read that Aikar [19:30] Aikar: its how i use to disable logging, app is working but still logging [19:31] qbit_ has joined the channel [19:31] odie5533: okay. I do a lot of programming with Python's Twisted library and would like to start using Nodejs instead. One of the benefits of using Python is that everything is easily portable. Would be very nice if/when Nodejs reaches that point as well. [19:32] cconstantine: odie5533, is that why you're asking about windows? I honestly see no reason to run node on windows but I"m fairly biased. [19:32] Aikar: `3rdEden: even with just io.set('log level', 0); it still logs [19:32] Aikar: >>> node app.js info - socket.io started info - unable to listen to port `843` as your node.js instance does not have root privileges. the flash policy file requests will only be served inline over the supplied http server. inline serving is slower than a dedicated server instance. [19:32] markwubben has joined the channel [19:33] mjr_ has joined the channel [19:33] `3rdEden: Aikar only 1 or 2 lines slips through before it's silenced [19:33] Aikar: hmm [19:33] `3rdEden: all other logs are killed [19:33] `3rdEden: also, if you see that line [19:33] odie5533: cconstantine: for one, all my application development is done on windows. So it'd be nice to not have to use vmware for development and testing. But I prefer running my applications on Windows as well, with the knowledge that when I upload them to a server they will work there as well. [19:33] `3rdEden: you should more your set loglevel above your set transports Aikar [19:34] mjr_ has joined the channel [19:34] Aikar: ah k [19:34] kavla has left the channel [19:34] Aikar: so no way to kill the socket.io started... ? :/ [19:34] kavla has joined the channel [19:34] `3rdEden: Aikar yes there is [19:34] `3rdEden: in 0.7.8 ;D [19:34] Aikar: where is 0.7.8! [19:34] hlindset has joined the channel [19:35] matomesc has joined the channel [19:35] `3rdEden: Aikar waiting in the github for it's release :) [19:35] cconstantine: odie5533, the reason you develop in linux is so that when you upload them to the server you know it'll work (because the server's linux too) ;) I'm not saying your wrong, I've just done a fair amount of development on windows and it has never been a good experience for me. [19:35] ale: How can I create a multichat app? [19:36] __directory: brianc: around? [19:36] odie5533: cconstantine: then you're doing it wrong! :D [19:36] odie5533: cconstantine: You need more Python in your life. [19:37] cconstantine: odie5533, oh god, python in windows? been there, hated it. [19:37] pyrony has joined the channel [19:37] __tosh has joined the channel [19:37] fille: i really need to stop create new github repository on my desktop :( [19:38] smathy has joined the channel [19:38] cconstantine: odie5533, the problem is windows. for dev/server things it's a joke os IMHO [19:38] cconstantine: s/joke/toy [19:38] fille: question am i allowed to use node in front of my project [19:38] __directory: allowed? [19:39] odie5533: erm... it's all I use. Python and Twisted, everything works always. Never had a problem. [19:39] Aikar: `3rdEden: this works too :P require('socket.io/lib/logger').prototype.log = function(){}; [19:39] isaacs: Does anyone know if node sends SNI stuff in its tls connections? (ryah pquerna) [19:39] euforic has joined the channel [19:39] `3rdEden: Aikar yup, [19:40] fille: creating a web mudclient based on node.js and want to name the git hub repository node-mud [19:40] Aikar: gonna leave a todo to remove it when ever socket.io gets updated to 0.7.8 [19:40] odie5533: Quuick! everyone register node-mud on github! [19:40] cconstantine: odie5533, cool. If it's working for you, then don't stop :) [19:40] fille: sorry but i already took mud-mume [19:40] fille: node-mume i mean [19:41] odie5533: cconstantine: What languages do you develop in? [19:41] nerdfiles1 has joined the channel [19:41] cconstantine: odie5533, including my windows time: C/C++, python. now python, ruby, javascript, bash [19:41] nerdfiles1 has left the channel [19:42] _root_ has joined the channel [19:42] odie5533: cconstantine: Sometimes I do GUI programming, too. And if I don't build it with Windows as my intended platform then I will have 0 userbase. [19:42] Aiden has joined the channel [19:43] cconstantine: odie5533, yup. thats why I was in windows before. so glad to not be doing desktop apps anymore [19:43] mjr__ has joined the channel [19:43] Draggor: Here's to hoping there's a nice union of node & QT [19:44] odie5533: hah ^ [19:44] kavla: yuck, that sounds terrible [19:44] odie5533: The union of Python and QT was/is brutal =/ [19:44] Cromulent has joined the channel [19:44] ekryski has joined the channel [19:44] odie5533: Licensing problems and the creation of two sets of bindings to add the letter L to the GPL license. [19:47] d0na1d has joined the channel [19:49] ph^ has joined the channel [19:50] toxico has joined the channel [19:51] ryah: isaacs: not yet [19:51] isaacs: ryah: what's necessary to make that happen? [19:51] ryah: isaacs: 0.5.3 added the initial support [19:51] isaacs: ah, ok [19:51] ryah: isaacs: bother indutny [19:51] toxico` has joined the channel [19:51] ryah: he's going to be finishing up support for it [19:51] isaacs: awesome, thanks [19:51] cafesofie has joined the channel [19:51] indutny: isaacs: em... ? [19:52] isaacs: with that, i think we'll have an npm that enterprises can feel safe and fuzzy about [19:52] isaacs: indutny: SNI in the node https client [19:52] indutny: ah [19:52] indutny: yeah [19:52] isaacs: indutny: the npm registry sits behind a vhost proxy on iriscouch, so the cert that gets served up has to be selected based on the hostname [19:52] indutny: it's in 0.5.3, but JS API is still not clear for us /cc pquerna ryah [19:53] indutny: isaacs: you can do this with current version of SNI [19:53] indutny: one second [19:53] indutny: I'll show you [19:53] balupton has joined the channel [19:53] indutny: isaacs: https://github.com/joyent/node/issues/1420 [19:53] isaacs: indutny: is it possibel in 0.4? [19:53] isaacs: or no [19:53] toxico`` has joined the channel [19:53] indutny: isaacs: no [19:53] isaacs: ok [19:54] guest_326 has joined the channel [19:54] indutny: but it can be ported if ryah will suggest doing so [19:56] ryah: indutny, isaacs: not going in v0.4 [19:57] indutny: :) [19:57] ryah: it's these new exciting features that get people to upgrade to the new code :) [19:57] indutny: good part is that it's working with default openssl lib [19:57] indutny: hahahaha [19:57] indutny: ryah++ [19:57] catb0t: ryah now has 1 beer [19:59] madsleejensen1 has joined the channel [19:59] mikelietz has joined the channel [19:59] isaacs: ok, so, in teh meantime, i guess i can try to get JasonSmith to set up a separate IP for the registry until we're ready to drop 0.4 support, or just rely on fingerprints and only do the full-fledged ssl dance in 0.5 [20:00] patcoll has joined the channel [20:01] mikelietz has left the channel [20:01] indutny: isaacs: hm... you can create simple nginx proxy [20:01] indutny: in Amazon EC2 [20:03] indutny: ping should be very small [20:03] indutny: and you'll get dedicated IP address [20:03] indutny: or host database yourself :) [20:03] indutny: hehe [20:04] isaacs: indutny: there is no way that i'd ever be as good a couch host as iriscouch. *maybe* if i made it my full-time job, but since Jason already made it his, that's a bit silly. [20:04] indutny: yeah, I know. just kidding [20:04] kimico has joined the channel [20:04] indutny: Jason is a man [20:05] blup has joined the channel [20:06] harthur has joined the channel [20:06] cafesofie has joined the channel [20:06] juli0w has joined the channel [20:10] puffpio has joined the channel [20:10] brez_ has joined the channel [20:12] towski has joined the channel [20:12] tristanseifert has joined the channel [20:12] mhauri has joined the channel [20:12] febits has joined the channel [20:15] pixel131 has joined the channel [20:16] greg has joined the channel [20:17] pixel13 has joined the channel [20:17] shinuza has joined the channel [20:17] ale: hey guys! [20:17] ryah: ale: hey [20:18] ale: Just a quick question [20:18] ryah: ale: shoot! [20:18] ale: If I use express then, I'll need to create an HTTP server, therefore if I have 99999999 connections, everything will collapse..am I right? [20:18] ale: The thing is that I want to get the url params, and I found out that the only way is with expressjs, am I right? [20:19] ryah: it's not possible to have 99999999 connections, no [20:19] ale: but with node and socketio yes! [20:19] heavysixer has joined the channel [20:19] ryah: you can get at them by manually parsing req.url [20:20] bshumate has joined the channel [20:20] ryah: but probably you should use express [20:20] jerrysv: bricks [20:20] toxico has joined the channel [20:20] ale: but if I use express I won't be able to have that many connections as if I don't use it! [20:20] jerrysv: ale: http://bricksjs.com/documentation.html [20:20] gazumps has joined the channel [20:21] guillermo has joined the channel [20:21] ale: jerrysv: does it use http? [20:21] jerrysv: ale: yes [20:21] indutny: ale: 10^8 -1 is too much even for node.js [20:21] ale: then how does facebook do to manage the connections? [20:22] indutny: horizontal scaling, I suppose? [20:22] tjholowaychuk: ale huh ? lol what does express have to do with limiting connections [20:22] indutny: running multiple instances [20:22] indutny: lets assume 1 connection = 1 kb memory [20:22] indutny: 10 ^ 11 bytes of memory [20:22] ale: because express uses http and http creates threads [20:22] tjholowaychuk: lol [20:22] jerrysv: um, what? [20:22] tjholowaychuk: ACTION headdesk [20:22] ale: and thread are limited [20:23] ale: maybe I have this concept wrong? [20:23] jerrysv: ale: um, i think you may have this concept wrong :) [20:23] ryah: ale: the http module doesnt use threads [20:23] jerrysv: i can understand people not wanting to use express for various reasons, but this is the first time i've heard that one :) [20:23] adrianolaru has joined the channel [20:24] ale: jerrysv: you know, every day you learn something new haha [20:24] indutny: 10^8 requests [20:24] ale: but when I call express I createServer() [20:24] tjholowaychuk: haha yeah that's up there with other bad assumptions [20:24] indutny: ryah: lets keep on going with optimizations [20:24] ale: so what should I do? [20:24] ale: indutny: 10⁸ requests per second? [20:24] jerrysv: ale: that just sets up the http socket and adds some events [20:25] ale: jerrysv: so then I won't be limited by using expressjs? What's the difference with bricks? [20:26] jerrysv: ale: it doesn't use connect, has multiple levels of routing, adds some nice event handling for the routes and errors [20:26] ale: jerrysv: so in some way it's much better [20:26] AvianFlu_ has joined the channel [20:27] ale: Ok, I'll analyze it! Thanks to all of you guys! [20:27] jerrysv: it's different -- different needs, different requirements [20:27] madsleejensen1: Hi all: im trying to do a simple request to a UDP bittorrent tracker. The protocol is http://www.bittorrent.org/beps/bep_0015.html#udp-tracker-protocol and my implementation is http://pastebin.com/TtTzUrv0 it sends the package fine but i never receive any response (i use Wireshark to verify that no response is ever sent to me) Do you have any ideas on whats wrong ? [20:27] robhawkes has joined the channel [20:28] Dreamer3 has joined the channel [20:28] tjholowaychuk: ale different != better. try things, see what you like, crack open the code etc [20:28] MooGoo has joined the channel [20:30] mscdex: madsleejensen1: node version? [20:30] markbao has joined the channel [20:31] madsleejensen1: 5.0 [20:31] madsleejensen1: mscdex: v0.5.0-pre [20:32] ruprict has joined the channel [20:32] madsleejensen1: mscdex: i tried using the writeInt32 methods aswell but there's no writeInt64 so i shifted the values myself [20:32] brez_ has joined the channel [20:33] kimico has joined the channel [20:33] mbrevoort has joined the channel [20:34] mscdex: madsleejensen1: are you requesting a connection id first? [20:34] bbrandon has joined the channel [20:34] rfay has joined the channel [20:35] madsleejensen1: mscdex: yes that is what im trying to do [20:35] Wa has joined the channel [20:36] heavysixer has joined the channel [20:36] mscdex: madsleejensen1: bit operators only work on 32-bit integers in javascript [20:36] dambalah has joined the channel [20:38] madsleejensen1: mscdex: hmm really ? then how would i send the 64 bit value ? [20:38] pifantastic_ has joined the channel [20:39] mscdex: madsleejensen1: you'll probably have to use a bigint module of some kind [20:39] arnee has joined the channel [20:39] qbert_ has joined the channel [20:39] mscdex: either pure js or c++ addon [20:39] qbert_: so there is no sleep() function ? [20:39] mscdex: qbert_: no [20:40] xippi has joined the channel [20:41] mscdex: madsleejensen1: for performance you might try: https://github.com/substack/node-bigint [20:41] mscdex: madsleejensen1: that one even supports writing to a buffer while specifying endianness [20:41] bnoordhuis: madsleejensen1: also, javascript numbers have < 64 bits precision [20:41] bnoordhuis: and what mscdex said [20:43] parshap has joined the channel [20:43] japj: I didnt know about the 64bit javascript int support until I saw the nodecommit pannel video today [20:43] japj: I'm glad the videos are finally showing up on blip.tv [20:43] theCode has joined the channel [20:44] ericmuyser has joined the channel [20:44] kriszyp has joined the channel [20:44] madsleejensen1: mscdex: alright i will have a look thanks! :) but couldnt i just do a var part1 = connectionId << 32; part2 = connectionId >> 32; [20:45] bnoordhuis: v8: 0xffffffff << 32 [20:45] bnoordhuis: aw, v8bot's dead? [20:45] SubStack: .. Math.pow(2,32) << 32 [20:45] mscdex: .. 0xffffffff << 32 [20:45] bnoordhuis: .. 0xffffffff << 32 [20:45] mscdex: heh [20:45] mscdex: all gone! [20:45] catb0t: Error: Timeout [20:45] mscdex: haha [20:46] mscdex: didn't know it was that computationally intensive ;-) [20:46] catb0t: Error: Timeout [20:46] juli0w has joined the channel [20:47] technoweenie has joined the channel [20:47] technoweenie has joined the channel [20:48] Cromulent has joined the channel [20:49] smathy: Maurice just pushed his 0.2 beta (rewrite) of zappa: https://github.com/mauricemach/zappa Looks very cool as far as node/coffee/express/socket.io/sammy frameworks go. [20:49] Determinist has joined the channel [20:51] dtan has joined the channel [20:52] adrianmg has joined the channel [20:52] hwinkel has joined the channel [20:55] csm_ has joined the channel [20:59] lightcap has joined the channel [20:59] samsonjs has joined the channel [20:59] AvianFlu has joined the channel [21:00] MaSch has joined the channel [21:00] _bat: smathy: pretty cool! [21:01] smathy: Yeah. [21:01] jerrysv: .. 1+1 [21:01] catb0t: 2 [21:02] jerrysv: works for me [21:02] mscdex: .. 0xffffffff << 32 [21:02] catb0t: -1 [21:02] mscdex: heh [21:02] ippa has left the channel [21:02] jerrysv: poor catb0t [21:03] catb0t: poor ending though [21:03] jerrysv: catb0t++ [21:03] tuhoojabotti: catb0t is alive [21:03] catb0t: alive in here started saying that node wasn't meant for searching because it's only for people who want to print the reason [21:03] jerrysv: catb0t is boring [21:03] catb0t: boring real fast is [21:03] `3rdEden has joined the channel [21:03] tuhoojabotti: yoda talk [21:03] jerrysv: catb0t is yoda [21:03] catb0t: yoda talk [21:04] tuhoojabotti: :D [21:05] puffpio has joined the channel [21:06] toxico has joined the channel [21:07] monokrome has joined the channel [21:08] euforic has left the channel [21:09] materialdesigner has joined the channel [21:09] stepheneb has joined the channel [21:09] stepheneb has joined the channel [21:10] ksheurs has joined the channel [21:10] stelleg has joined the channel [21:10] zemm has joined the channel [21:11] ksheurs has joined the channel [21:11] kepakiano has joined the channel [21:12] zeade has joined the channel [21:12] kepakiano has left the channel [21:13] materialdesigner: anyone have experience with using node-oauth or connect-auth and oauth 2.0? [21:13] materialdesigner: I can't seem to figure out how to actually *do* anything [21:13] bnoguchi has joined the channel [21:14] bnoguchi: materialdesigner: have you checked out https://github.com/bnoguchi/everyauth ? [21:15] materialdesigner: ah, no I hadn't [21:15] brianc: __directory: hi [21:15] materialdesigner: I see you are the dev [21:15] materialdesigner: ACTION snuggles up to bnoguchi [21:15] bnoguchi: materialdesigner: yea. check it out; lots of documentation. drop me a line on github if you need help with it. [21:16] materialdesigner: k sounds good [21:16] hlindset has joined the channel [21:16] brianc: +1 for everyauth [21:16] brianc: it's good stuff [21:16] ekryski: ditto ^ [21:16] materialdesigner: it allows you to connect to other oauth2 providers, right? [21:17] materialdesigner: not just the ones listed? [21:17] bnoguchi: materialdesigner: yes [21:17] materialdesigner: k :) [21:17] ruprict: So, as long as were on the subject of everyauth *cough* [21:17] brianc: __directory: I should be around for the next 30 min or so. if i'm gone, drop me a line @ brian.m.carlson@gmail.com [21:18] ruprict: I have (what is likely a silly) question about websockets and auth. Using the classic node_chat example, what if I wanted to persist each message inside a User document? Think user.messages [21:20] Aiden has joined the channel [21:20] ruprict: How can I get the user on a socket.message? (Or, can I?) [21:20] bnoguchi: ruprict: You just broadcast the user over a channel if using socket.io [21:21] hlindset has joined the channel [21:21] adambrault has joined the channel [21:22] mbrevoort has joined the channel [21:22] qbert_: ohhh everyauth, good find [21:22] ruprict: bnoguchi: Um….so, after user autheticates? I can broadcast on the server to a channel (that doesn't go to the client, obv)? [21:23] qbert_: npm should have some discovery something or other on its website :) [21:24] HAITI has joined the channel [21:24] isaacs: qbert_: npm search on the cli works well [21:24] isaacs: qbert_: or search.npmjs.org [21:29] qbert_: yes search does work well [21:29] threecreepio has joined the channel [21:29] ruprict: bnoguchi: I don't think I explained myself very well. Authenticated user sends chat message, client sends message over web socket. I want to find that user in the db and add the message to that user's pastMessages (or whatever). Then broadcast the message to everyone else. [21:30] loveshine has joined the channel [21:30] bnoguchi: ruprict: Just add a senderId to each message object. [21:31] mikeal has joined the channel [21:31] bnoguchi: ruprict: Then on the server, you can find the user by id, append the message to the user's messages, and then broadcast [21:31] ryah: http://nodejs.org/meetup/ [21:32] hlindset has joined the channel [21:32] ruprict: bnoguchi: Thx…btw mongoose_auth and everyauth are both fantastic…*buys bnoguchi a beer* [21:32] berasa has joined the channel [21:33] skohorn has joined the channel [21:33] puffpio has joined the channel [21:33] bnoguchi: ruprict: Alternatively, you can send a setUserId message as the first socket payload. Then, for each incoming message, lookup the user by the associated socket userId [21:33] japj: ryah: any chance it'll be streamed on the internet? [21:33] ryah: people in SF, you should try to make it to this meetup http://nodejs.org/meetup [21:33] ctide: bnoguchi++ [21:33] ryah: it's going to be very fun! [21:33] catb0t: bnoguchi now has 1 beer [21:33] stepheneb has joined the channel [21:33] ryah: japj: we'll make videos available afterwards [21:34] japj: ryah: thanks! [21:34] pquerna: ryah: going to vegas for defcon :-/ [21:35] HenrikJoreteg has joined the channel [21:36] bnoguchi has left the channel [21:37] skampler: is there any official effort towards a distributed node architecture? [21:37] skampler: something in the lines of fault-tolerant way of running thousands of node processes on different machines and passing messages around [21:38] ngs has joined the channel [21:38] AvianFlu: skampler: http://github.com/hookio/hook.io [21:39] AvianFlu: maybe not 'thousands' just yet, but 'an effort' in that direction for sure [21:40] skampler: c00l [21:41] loveshine_ has joined the channel [21:41] wookiehangover has joined the channel [21:41] mikeal: oh man, Defcon [21:41] mikeal: i haven't been in like 10 years [21:41] CIA-65: node: 03Ben Noordhuis 07master * r38f948a 10/ (Makefile src/node.cc src/node.js tools/test.py): [21:41] CIA-65: node: http: make http2 the default, legacy backend is available with --use-http1 [21:41] CIA-65: node: Fixes #1441. - https://github.com/joyent/node/commit/38f948a373aa3dc8c46628bbd53edf07b6f16ea7 [21:42] michaelhartau has joined the channel [21:42] unlink has joined the channel [21:42] pjacobs has joined the channel [21:42] kartmetal_ has joined the channel [21:42] kartmetal_ has joined the channel [21:46] kartmetal__ has joined the channel [21:46] kartmetal__ has joined the channel [21:47] rfay has joined the channel [21:47] Anton_: how can i identify the reason of segmentation fault? [21:48] Anton_: it's just node is crashing silently and only "segmentation fault" line is there [21:48] Anton_: is there some debug switches or log files? [21:49] TheJH: Anton_, try "node --help" [21:49] bnoordhuis: Anton_: `./configure --debug` and run node_g through gdb [21:49] bnoordhuis: Anton_: or try hooking it up to gdb right now [21:49] bnoordhuis: but you won't have symbolic names [21:50] eignerchris has joined the channel [21:50] ryah: Anton_: what os? [21:50] Anton_: debian 5 [21:50] dherman has joined the channel [21:50] ryah: Anton_: gdb --args ./node script.js [21:50] ryah: Anton_: get us a stacktrace [21:50] ryah: or coredump [21:50] Anton_: thanks [21:53] dguttman_ has joined the channel [21:54] losing has joined the channel [21:54] Drakonite has joined the channel [21:55] mike5w3c has joined the channel [21:55] robertfw has joined the channel [21:57] devongovett has joined the channel [21:58] kevindente has joined the channel [21:59] DrPizza: ryah: Is there any quick and easy way to make the testing more aggressive? [22:00] dherman has joined the channel [22:00] DrPizza: ryah: like at the moment the windows builds IIRC "pass" the openssl tests, because the openssl tests bail early if openssl isn't built in [22:00] mikeal has joined the channel [22:00] gazumps has joined the channel [22:00] DrPizza: when surely that should really be a failure [22:01] cjm has joined the channel [22:01] balaa has joined the channel [22:02] blup has joined the channel [22:06] iFire has joined the channel [22:07] Xano has joined the channel [22:07] hij1nx has joined the channel [22:08] pylon__ has joined the channel [22:09] fairwinds has joined the channel [22:10] hebz0rl has joined the channel [22:12] davidbanham has joined the channel [22:14] gozala has joined the channel [22:15] eee_c has joined the channel [22:15] ryah: DrPizza: build with ssl [22:15] ekryski1 has joined the channel [22:16] ryah: DrPizza: there aren't too many tests that allow you to get by that easily [22:16] amscotti has joined the channel [22:16] DrPizza: I thought the win32 builds omitted ssl. [22:16] arafatm has joined the channel [22:16] DrPizza: (I actually seem to have it more or less working in my fork, but still needs some tweaking to the builds) [22:18] zaach has joined the channel [22:18] yhahn has left the channel [22:18] zaach has joined the channel [22:18] DrPizza: ryah: would there be interest in patches to make node build in MSVC++? I notice that libuv does, but node itself does not. [22:19] pita has joined the channel [22:23] felixge has joined the channel [22:23] felixge has joined the channel [22:24] bnoordhuis: DrPizza: we're planning on switching to gyp [22:24] DrPizza: ugh [22:24] DrPizza: gyp produces really crappy project files [22:24] CIA-65: libuv: 03Bert Belder 07master * rdbaddc4 10/ src/win/pipe.c : Windows: avoid errorneous freeing of pipe handle - https://github.com/joyent/libuv/commit/dbaddc4423d61fa16cca299650f8378cffba9cce [22:25] DrPizza: when you open them up in 2010 they produce a variety of semi-spurious errors, and they do some things (like custom build rules) in a way that's just plain wonky [22:25] pixel13 has joined the channel [22:26] aconbere has joined the channel [22:26] Nuck has joined the channel [22:27] patrickgamer has joined the channel [22:29] pixel13 has left the channel [22:29] patrickgamer has left the channel [22:30] zemanel has joined the channel [22:30] TheJH: guys, can you come up with an explaination for these results? why is the "minified" test FASTER than "plain slice"? http://jsperf.com/minify-index-perfimpact [22:30] TheJH: this is crazy [22:31] gozala has joined the channel [22:31] espacia has joined the channel [22:33] pixel131 has joined the channel [22:33] TheJH: gonna cross-post this to ##javascript [22:33] heavysixer has joined the channel [22:35] CIA-65: node: 03Bert Belder 07master * r6efc86d 10/ deps/uv/src/win/pipe.c : Upgrade libuv to dbaddc4423d61fa16cca299650f8378cffba9cce - https://github.com/joyent/node/commit/6efc86df6a91fca05791812bbbfc4d265c1cb387 [22:38] bnoordhuis: TheJH: b[a]() is slower here [22:38] bnoordhuis: here = node master [22:38] chapel: its barely different [22:39] chapel: 1k difference out of 9million [22:39] bnoordhuis: TheJH: https://gist.github.com/2e142704b884e93965b3 [22:39] darshanshankar has joined the channel [22:40] zivester has joined the channel [22:41] TheJH: bnoordhuis, getting the same kind of results here... [22:41] TheJH: seems like the performance depends on the version of v8 [22:41] TheJH: I'll try it in firefox... [22:43] mjr_: pquerna: I forgot that there even was defcon. But aren't you a bit too old for that sort of thing these days? [22:43] TheJH: "minified" is 20% slower in firefox... mhm... [22:43] skm has joined the channel [22:43] brianseeders has joined the channel [22:44] piscisaureus: DrPizza: windows builds with openssl by default [22:44] colinclark has joined the channel [22:44] brianseeders has joined the channel [22:45] DrPizza: piscisaureus: it does? doesn't it get commented in wscript? [22:45] DrPizza: or opted out [22:45] DrPizza: or was I looking at an old version/ [22:45] DrPizza: it sure didn't [22:46] qbert_: ryah, http://www.meetup.com/noders/ :) [22:46] ryah: qbert_: nice :) [22:47] smus has joined the channel [22:47] DrPizza: piscisaureus: or at least [22:47] DrPizza: piscisaureus: it told me it couldn't find it and that I had to exclude it [22:48] DrPizza: piscisaureus: is there something I need to do to make it find it? [22:48] DrPizza: piscisaureus: it told me "Could not autodetect OpenSSL support" [22:49] DrPizza: oh it's looking for it where, in the parent? [22:50] bnoordhuis: DrPizza: i think in ../openssl [22:50] odyniec_ has joined the channel [22:50] DrPizza: hmm why there and not in deps? [22:51] DrPizza: I guess on unix you use the system libssl instead of static linking [22:53] qbert_: sucks, I'll be in SF a day after the official node meetup :( [22:53] kriszyp has joined the channel [22:54] HenrikJoreteg has joined the channel [22:54] monokrome has joined the channel [22:54] skm has joined the channel [22:55] DTrejo: qbert_: :| [22:56] feychenie has joined the channel [22:57] HenrikJoreteg has joined the channel [22:58] hlindset has joined the channel [22:58] hlindset has joined the channel [22:58] DrPizza: piscisaureus: so if I put it in ../openssl and do a mingw build it'll be able to find the libraries it needs after all/ [22:58] traph has joined the channel [22:59] DrPizza: is there some reason for not distributing binaries like you do for pthreads? [23:00] gg777 has joined the channel [23:00] arafatm has left the channel [23:01] HenrikJoreteg has joined the channel [23:01] tristanseifert has joined the channel [23:01] mikeal has joined the channel [23:08] scott_gonzalez has joined the channel [23:12] puffpio has joined the channel [23:15] aoberoi has joined the channel [23:16] al3xnull has joined the channel [23:18] pixel131 has left the channel [23:18] pixel13 has joined the channel [23:18] pixel13 has left the channel [23:21] theCole_ has joined the channel [23:24] ericmuyser has joined the channel [23:26] Croms has joined the channel [23:27] ifandelse has joined the channel [23:28] mikeal has joined the channel [23:28] ifandelse: hi everyone. I'm curious as to which debugging tool you use for node. I've been using node-inspector and have been pretty happy with it. Anything else I should be taking a look at? [23:28] losing has joined the channel [23:29] mikeal has joined the channel [23:29] kevindente: ifandelse: Eclipse with V8 debugging is pretty decent [23:30] hydrozen has joined the channel [23:30] tjholowaychuk: ifandelse node's built-in debugger $ node debug script [23:31] migimunz has joined the channel [23:32] materialdesigner: how do i specify other parameters to give to the authPath in everyauth? [23:33] materialdesigner: i.e. It's constructing the URL properly but leaving off a &type=web_server [23:33] just_monkey_code has joined the channel [23:33] amscotti has left the channel [23:33] smathy has left the channel [23:33] smathy has joined the channel [23:36] DevX101 has joined the channel [23:36] mikeal has joined the channel [23:36] mapleman has joined the channel [23:37] just_monkey_code: New to node and am wondering how to redirect stderr to a file. [23:38] skampler: just_monkey_code: node prog.js 2>file [23:40] tar_ has joined the channel [23:40] ajsie: anyone nows where I can watch the node.js 2011 conf videos? [23:40] just_monkey_code: Is there a way to do this from within the module itself? I know that you can create child processes and give them what ever stderr you would like, but is there a way that a module can control it's own stderr? [23:42] CIA-65: libuv: 03Bert Belder 07master * rec14396 10/ src/win/process.c : Windows: fix bug in process arguments buffer size computation - https://github.com/joyent/libuv/commit/ec143961d135adb3f49f5a2322774ef43e2520b9 [23:42] replore has joined the channel [23:43] just_monkey_code: An example may be a server which (upon startup) reads a settings file and depending on the value of something defined in that file sends stderr either to the console or to a file [23:43] odie5533: Is there a page or file that discusses the Windows port progress? [23:43] simenbrekken has joined the channel [23:44] AvianFlu: just_monkey_code: process.stderr.on('data') maybe [23:44] just_monkey_code: Thank you. I'll give it a try. [23:45] DEVX_ has joined the channel [23:45] galaxywatcher has joined the channel [23:45] CIA-65: node: 03Bert Belder 07master * r09bb1d6 10/ deps/uv/src/win/process.c : Upgrade libuv to ec143961d135adb3f49f5a2322774ef43e2520b9 - https://github.com/joyent/node/commit/09bb1d64dc774bc52f1b0dfd048715ad3243e550 [23:46] monokrome has joined the channel [23:51] arpunk has joined the channel [23:51] azeroth__ has joined the channel [23:51] rfay has joined the channel [23:51] just_monkey_code: no dice on process.stderr.on('data'). Per the docs the data event is only emitted on Readable streams ( the same issue with the pipe() method ). stderr is only a Writeable stream. [23:52] AvianFlu: oh yeah, cause it's not a child process [23:52] AvianFlu: you could pipe it to a readable stream [23:52] AvianFlu: and then use that stream's data event to send it wherever [23:53] just_monkey_code: That's a good idea. Let's give that a shot :) Thanks [23:53] AvianFlu: but really, just send it somewhere else before you write it to stderr [23:54] mjr__ has joined the channel [23:56] stepheneb has joined the channel [23:56] kevindente has left the channel [23:58] losing has joined the channel [23:58] davidbanham has joined the channel [23:59] hlindset has joined the channel [23:59] hlindset has joined the channel [23:59] TheEmpath has joined the channel [23:59] isaacs: hij1nx: hey