00:00:00 * ircretary quit (Remote host closed the connection) 00:00:08 * ircretary joined 00:00:24 * spion joined 00:00:34 I'm really rusty with this stuff! 00:01:05 But basically it simulates perfect decision-making for an agent buying stuff 00:01:25 only major problem is that this doesn't consider the fact that there's an overhead per-vendor 00:01:43 you could farm out the AI part of it to use voxeljs 00:01:45 *UI 00:01:57 so you would just plop down voxel bricks 00:02:03 for the servers 00:02:04 and cables 00:02:06 possibly 00:02:23 but I'm really thinking of using a cli/repl for the ux 00:02:29 at this point 00:02:41 we'll see though 00:02:47 at this point that's all there is 00:02:59 anyone base .gitignore parser ? 00:04:32 .npmignore is more or less compatible with .gitignore right? 00:04:35 maybe use that gozala 00:04:56 jesusabdullah: yeah I was actually trying to find that code 00:05:02 but can't really find it yet 00:05:24 super super crude, but works for my usecase: https://github.com/marcello3d/node-safestart 00:06:10 * yorick quit (Remote host closed the connection) 00:09:49 * kenperkins joined 00:10:22 * kenperkins quit (Remote host closed the connection) 00:11:04 * kenperkins joined 00:16:06 * CryptoQuick joined 00:18:17 crazy idea: using git and rss to stream together a fundamentally distributed social network 00:18:35 appropriating what already exists with rss 00:18:51 but doing the network bootstrapping and traversals 00:20:26 * CryptoQuick quit (Ping timeout: 252 seconds) 00:21:15 * CryptoQuick joined 00:22:42 * AvianFlu joined 00:23:08 im hacking on a thing right now that will archive youtube documentaries that i want to watch on google storage so i have a copy foreva 00:24:00 * thl0 joined 00:25:54 * CryptoQuick quit (Ping timeout: 264 seconds) 00:25:56 nice 00:33:04 gozala: https://github.com/isaacs/fstream-ignore 00:33:05 oh maaaaan I have to try deriving the equations again 00:33:07 MOAR LATERS 00:33:08 I WANT SOME GOD DAMNED APPLE CIDER BITCH, GO PRESS YOUR APPLES INTO THE MAMMOGRAM UNTIL YOU SCREAM! 00:33:30 https://gist.github.com/jesusabdullah/5166336#file-test-js example use tho 00:34:12 * Domenic_ joined 00:38:00 * mikolalysenko quit (Ping timeout: 264 seconds) 00:38:53 * mikolalysenko joined 00:39:38 substack: related http://goo.gl/HFEbV 00:40:08 i dont agree with the article but anyway 00:42:59 * mikolalysenko quit (Ping timeout: 246 seconds) 00:45:06 I think RSS feeds are like talk radio 00:45:23 Very Polarizing. Some people never use them, others use them to a ridiculous degree 00:45:30 without a lot of middle ground. 00:56:10 * marcello3d quit (Remote host closed the connection) 01:05:27 * nk109 quit (Quit: Computer has gone to sleep.) 01:10:21 * Domenic__ joined 01:12:06 substack: have you had any issues with tape in 0.10? 01:12:17 I haven't used 0.10 yet 01:12:22 ah 01:12:23 cool 01:13:38 chrisdickinson: https://github.com/substack/tape/issues/21 01:14:01 mbalho: ah, i'm looking at a possible out of order thing 01:14:20 my tests for tar-parse work in 0.8.X, but skip a bunch of tests in 0.10.0 01:14:31 like, they run the test case but never run any assertions? 01:14:46 * Domenic__ quit (Ping timeout: 252 seconds) 01:15:27 yeah, it definitely looks like it's not waiting for t.end() for some reason 01:16:35 * ins0mnia quit (Ping timeout: 258 seconds) 01:21:16 * nk109 joined 01:22:09 * dominictarr joined 01:26:26 isaacs: is there a refactored moudle that you use to parse website, homepage, and repo from package.json? 01:32:09 * mikeal joined 01:33:24 thl0: do the source maps in coffeeify only show up in the final output when somebody does --debug? 01:33:35 anyways I don't really care that's what minifiers are for 01:34:35 substack: coffeeify itself has no notion of --debug, but browser-pack will only use them when --debug is on (via setting row.sourceFile) 01:35:00 ok sweet! 01:35:03 substack: browser-pack will also strip any existing mappings from file sources as to not confuse the browser ;) 01:35:18 I was more thinking what does browser-pack do if there are source maps in a file but debug is off 01:35:23 aha yep that's perfect 01:36:00 substack: so if sourceFile is set it will try to offset existing mappings before removing them 01:36:53 substack: if no original source is inlined in the mapping of the passed source, it willjust ignore the entire mapping and generate them from scratch 01:37:01 published 01:38:33 substack: nice! working on getting browser-pack to consume those properly 01:39:18 so excellent 01:43:20 dominictarr: not sure. maybe. 01:43:29 dominictarr: if not, patch welcome 01:43:45 isaacs: okay, I'll have a look 01:49:58 dominictarr, Raynos, juliangruber, ralphtheninja, hij1nx: https://github.com/rvagg/node-levelup/pull/102#issuecomment-14939884 01:50:11 pass in a 'db' option to swap out LevelDOWN 01:50:24 can use this as a base for your replacement: https://github.com/rvagg/node-abstract-leveldown 01:50:54 it passes all the basic LevelDOWN tests (for arguments and some state stuff), so you just need to add concrete implementations to do the actual work.. see the example for an in-memory store 01:51:08 .. which I plan to write next, a full in-memory replacement for LevelDOWN 01:51:12 rvagg: SWEET 01:59:28 * mikolalysenko joined 02:16:22 * tilgovi joined 02:18:22 http://substack.net/many_things 02:19:54 substack: so, process.nextTick fails in tape on 0.10.0. switching to setImmediate works, but that's yet-another-thing-to-stub in browserify, probably 02:20:55 crap 02:20:55 someone needs to write a blog post on "went to use setImmediate" 02:21:29 Raynos: the new version of https://github.com/dominictarr/readme will open the github page of a module 02:21:36 readme module --gh 02:21:49 oh man I am going to blog so much you have no idea 02:21:58 this blogging engine is SO EASY 02:25:31 * jibay joined 02:25:49 rvagg: nice! 02:26:52 WOOT i wrote a node module that streams youtube videos into google storage 02:27:18 so i can watch documentaries that have been DMCA'd MUAHAHA 02:27:48 mbalho: cool :) 02:27:58 mbalho: did you see my message the other day about data? 02:29:01 ralphtheninja: maybe? where did you ask it? 02:29:24 I actually have been thinking about a setImmediate blog post 02:30:32 mbalho: on #stackvm .. I was asking if you have any pointers to data, it could be any kind of statistics really, preferrably in json :) 02:30:57 oh must have missed that 02:31:11 mbalho: different coordinates on the planet, weather stuff, anything 02:32:04 * mikeal quit (Quit: Leaving.) 02:32:30 ralphtheninja: heres a buncha oakland stuff http://data.openoakland.org/ 02:32:44 ralphtheninja: also this meta site http://datacatalogs.org/ 02:34:44 mbalho: awesome, thanks! 02:35:32 * CoverSlide quit (Ping timeout: 248 seconds) 02:36:33 * CoverSlide joined 02:37:55 * thl0 quit (Remote host closed the connection) 02:48:23 * marcello3d joined 02:52:47 * marcello3d quit (Ping timeout: 252 seconds) 02:58:28 * CoverSlide quit (Ping timeout: 240 seconds) 03:02:14 * mikeal joined 03:03:36 * marcello3d joined 03:03:56 * CoverSlide joined 03:08:55 * CoverSlide quit (Ping timeout: 264 seconds) 03:11:19 https://github.com/maxogden/youtube2googlestorage 03:11:48 next step is to take https://github.com/maxogden/download-all-the-things and turn it into a module that gets all youtube links from a subreddit 03:11:55 * mikeal quit (Ping timeout: 264 seconds) 03:13:05 * dominictarr quit (Quit: dominictarr) 03:17:22 STREAMS! 03:17:31 aww 03:25:12 * jibay quit (Quit: Leaving) 03:29:31 https://gist.github.com/jesusabdullah/5167288 ;_; 03:29:35 noweb 03:29:37 \disappoint 03:29:41 ಠ_ಠ 03:29:42 I MEAN 03:29:46 you can see I've been texing 03:34:21 \section*{diappointments} 03:34:28 yes 03:34:43 \begin{itemize} 03:35:02 \item latex 03:35:06 \item too meta 03:35:09 \end{itemize} 03:36:54 substack: should i open that setImmediate/nextTick thing as an issue on tape? 03:38:46 * mikeal joined 03:39:42 if it breaks on 0.10 then sure 03:39:52 I'll add 0.10 to the travis tests 03:43:00 * shama quit (Ping timeout: 258 seconds) 03:49:08 can someone help me figure out what i'm doing wrong with this 0.10.0 readable stream implementation? https://gist.github.com/maxogden/8a05d3fa3615e7404c8e 03:49:12 (scroll down for error) 03:50:52 * mikeal quit (Quit: Leaving.) 03:58:06 nvm was piping object stream -> non object stream 03:58:13 that error could be way better 03:58:44 jesusabdullah: what are you trying to solve? 03:58:49 * mbalho forks node 03:59:16 if it is maximize utility * purchases 03:59:24 subject to purchases * price <= budget 03:59:30 then, that is just linear programming 03:59:39 and you can solve it in one line using numeric.js' solveLP 04:00:27 http://numericjs.com/numeric/documentation.html 04:00:34 scroll down to solveLP() 04:01:31 you can also do handle limited supplies of quantities 04:02:36 * mikeal joined 04:03:53 * mikeal quit (Client Quit) 04:08:39 mbalho: i've just been using through still in the hopes of avoiding upcoming weirdness :| 04:09:01 i like the new streams 04:09:15 i really like through's api. 04:09:35 haven't messed much with streams2, though 04:10:05 mbalho: any place in particular that streams2 shines? 04:11:05 chrisdickinson: its easier to implement a readable stream IMO, the base classes are nice 04:11:13 chrisdickinson: biggest difference is if youre writing, for example, an auth server 04:11:21 chrisdickinson: lets say authenticated video uploads 04:11:35 chrisdickinson: you have to do an async call to your session store before accepting the video stream upload 04:11:50 * chrisdickinson nods 04:11:50 chrisdickinson: before you had to pipe to a buffered stream but now you can just wait to read it 04:12:10 chrisdickinson: overall the semantics are very similar it just all seems a lot simpler to me 04:12:17 no more pausing, then 04:12:21 theoretically 04:12:51 (dominictarr explained it as "think of it as normal streams but in reverse", roughly, at cascadiajs) 04:14:18 i think the mental difference of going from not streaming to streaming is wayyyy bigger than going to streams1 to streams2 04:14:26 it seems more complex to me (but not wrong), but i'm mostly thinking of the <0.10 stream file where it just defined pipe and nothing else. 04:15:03 * chrisdickinson nods 04:15:36 streams1 and streams2 are the same basic concept with a switch in regards to who's in control of determining whether to send more data 04:15:40 x = new require('stream').Readable(); x._read = function(n) {}; otherThing.on('data' x.push) 04:15:53 _read gets called whenever something wants data from x 04:15:56 e.g. pipe 04:16:18 otherThing should probably be called xDataSource 04:16:48 i haven't really thought through all of the implications yet 04:16:55 but yeah, it definitely makes more sense 04:17:20 since we had a boolean signal exposed as to whether more data could be received or not, and it made things fuzzy on the other end of that signal 04:17:46 * defunctzombie_zz changed nick to defunctzombie 04:18:04 mbalho: streams1 is just a thing that reads infinitely then? 04:18:39 well, reads until it itself pauses, then waits for a drain event on itself to resume reading? 04:19:06 what might that mean for domnode-ish stuff? 04:22:57 chrisdickinson: hmm not sure yet 04:23:35 yeah, that's kind of where i'm struggling :| 04:24:19 i suppose the streams1 emulation makes the domnode stuff work as expected 04:26:23 insert-css is really not the right way to make reusable css for modules 04:27:47 defunctzombie: what is the right way? 04:28:05 substack: letting me get your css file for your widget if I want it 04:28:14 substack: and documenting the css classes used that I can style 04:28:17 defunctzombie: so the problem is 04:28:24 a lot of css is highly structural in nature 04:28:35 it needs to be exactly so or else the widget is completely unusable 04:28:41 substack: sure, that is fine, then your css needs to be minimal 04:28:48 substack: and also namespaced 04:28:59 and not inserted into the head for me 04:29:02 which just means that you're going to be encoding things that would normally be in css in your javascript 04:29:17 substack: I still think you should put them in css 04:29:26 and allow me to use your css files via tools 04:29:29 defunctzombie: insert-css does not force you to insert css 04:29:33 or not at all if I dont want 04:29:43 it could just be a function you call that inserts optionally 04:29:48 substack: I am saying widgets should not do this 04:29:51 that decision is up to the module consumers 04:29:52 i'm on the side that any css that is absolutely fundamental should be with the js 04:29:58 nor should they be encouraged to do it 04:30:30 you should not be inserting css into the head of the document within your widget 04:30:41 this approach is very convenient and I don't really get what the alternative to doing it this way is 04:30:47 this is like monkey patching global objects in JS 04:30:52 no, ideally you insert scoped css 04:30:58 to your element 04:30:59 you just scope the declarations and it works pretty good 04:31:13 http://html5doctor.com/the-scoped-attribute/ 04:31:24 * dguttman quit (Ping timeout: 264 seconds) 04:31:24 chrisdickinson: yes the shadow dom stuff will let us do that 04:31:27 but until then 04:31:29 * chrisdickinson nods 04:31:45 i'm not certain that i like the shadow dom in particular 04:31:47 don't like it, and don't think this is good at all 04:32:01 substack: I just use npm-css and allow people to opt in 04:32:09 element scoped css should solve the problem, largely 04:32:22 also, css variables 04:32:38 which are interesting in that they follow css rule specificity 04:32:49 step 1) get everyone to change their fundamental behavior is never my stype 04:33:02 style* 04:33:04 defunctzombie: I think npm-css adds too much involvement at the module consumer level 04:33:21 I would rather that consumers didn't need to care that widgets even have css 04:33:31 that isn't how css works 04:33:33 fundamentally 04:33:45 and sure, scoped will change that when we get there 04:33:58 but until then, I think this is an anti-pattern approach personally 04:33:59 another trick I really like 04:34:07 * mikeal joined 04:34:15 is just bundling html strings with