00:00:01  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:15:00  <Nexxy>substack, is there by chance a burrito tunnel to Portland?
00:19:24  <jesusabdullah>substack: https://gist.github.com/jesusabdullah/5178945 C:
00:22:46  * st_lukejoined
00:23:33  <defunctzombie>how to tell forbesl that no one cares about other module loaders haha
00:24:07  <thl0>substack: good news and bad news
00:24:26  <thl0>substack: good I got coffee script working with sourcemapped bundle
00:24:56  <thl0>substack: bad generating sourcemaps for large files is super slow, I think this is the culprit: https://github.com/thlorenz/inline-source-map/blob/a80d11c86160fa8948ed1e4900825c961c895620/index.js#L59-L66
00:25:39  <thl0>could anyone give me a hint how to make this perform better (is it possibly the GC going crazy with all the location objects I create)?
00:26:26  <thl0>oh shoot just saw that this is called on every iteration: newlinesIn(source) that's gonna be expensive ;)
00:27:48  <thl0>that was it - so substack: only good news left :)
00:32:03  <jesusabdullah>substack: I mean https://gist.github.com/jesusabdullah/5178955
00:41:48  <defunctzombie>http://shtylman.com/post/node-error-logging/
00:41:55  <defunctzombie>anyone wanna give me some feedback ^ ?
00:46:03  <jesusabdullah>My concerns with logging have to do more with how to pretty-print them than anything
00:46:15  <jesusabdullah>also sometimes the errors aren't errors
00:46:22  <jesusabdullah>which is bullshit but it happens
00:46:40  <jesusabdullah>and on some level I'm "over" logging frameworks
00:46:44  <jesusabdullah>but beyond that..!
00:47:12  <defunctzombie>I usually don't care about pretty printing them
00:47:16  <defunctzombie>I care about recording them
00:47:18  <defunctzombie>and alerting
00:47:20  <defunctzombie>and meta info
00:47:33  <defunctzombie>pretty printing is useless cause my services are not being watching by me 24/7
00:47:45  <defunctzombie>I have all the pretty printing I want with a stacktrace when debugging locally
00:48:01  <jesusabdullah>but printing an error in a way that doesn't show the stacktrace or the metadata or otherwise breaks is useless too
00:48:09  <jesusabdullah>which happens!
00:48:37  <defunctzombie>yea, book just prints the message and stacktrace to stdout
00:48:40  <defunctzombie>(by default)
00:48:46  <defunctzombie>but you can do whatever you want
00:48:46  * thl0quit (Remote host closed the connection)
00:48:59  * douglaslassancejoined
00:49:41  * douglaslassancechanged nick to asterokid
00:58:12  * mikealjoined
00:59:29  * mikealquit (Client Quit)
01:02:26  * mikealjoined
01:02:44  * mikealquit (Client Quit)
01:05:33  * Domenic_quit (Ping timeout: 276 seconds)
01:06:56  * ins0mniaquit (Ping timeout: 255 seconds)
01:33:00  * mikealjoined
01:41:01  * AvianFluquit (Remote host closed the connection)
01:41:23  * mikealquit (Ping timeout: 245 seconds)
01:45:42  <st_luke>what about IO
01:47:19  <defunctzombie>st_luke: I got the irc bouncer setup btw
01:47:33  <st_luke>did you end up doing it on that super cheap vps host?
01:47:37  <defunctzombie>yea
01:47:46  <defunctzombie>just got a separate instance for it
01:48:09  <defunctzombie>there are some online services that do it for free apparently
01:48:17  <defunctzombie>but none of them looked like something I wanted to use
01:48:22  <defunctzombie>so I just went with the server
01:49:03  <st_luke>yea paas for irc is a bit much
02:00:49  * thl0joined
02:07:22  * nichola__joined
02:09:19  * mikealjoined
02:10:24  * nicholasfquit (Ping timeout: 264 seconds)
02:31:46  <thl0>substack: check it out: http://thlorenz.github.com/WebGLCraft/
02:32:06  <thl0>substack: browserified with CoffeeScript sourcemaps preserved
02:33:21  <substack>no pointerlock?
02:33:53  <thl0>substack: pointerlock? what do you mean?
02:35:21  <thl0>substack: I know it's not voxeljs, but I needed something in CoffeeScript and this was the best I could find ;)
02:36:00  <substack>ah that's why
02:42:24  * ircretaryquit (Ping timeout: 276 seconds)
02:47:43  * ircretaryjoined
03:00:31  * asterokidquit (Quit: Linkinus - http://linkinus.com)
03:22:24  * No9quit (Ping timeout: 240 seconds)
03:22:34  <defunctzombie>thl0: is that supposed to have sourcemaps on the page?
03:22:49  <thl0>defunctzombie: yes
03:22:53  <defunctzombie>nice it does
03:22:55  <defunctzombie>amazing
03:23:03  <thl0>defunctzombie: works for you?
03:23:06  <defunctzombie>yea
03:23:12  <defunctzombie>I had to turn it on in the browser
03:23:25  <thl0>defunctzombie: cool eh? time to make that pull request to substack
03:23:53  <thl0>defunctzombie: browserpack had to change only a tiny bit - I'm doing most of the work in a module I created
03:23:59  * No9joined
03:24:01  <defunctzombie>cool
03:24:15  <defunctzombie>how is the perf for sourcemaps?
03:24:24  <defunctzombie>for js or coffee?
03:24:31  <thl0>defunctzombie: good enough as far as I can tell
03:24:54  <defunctzombie>cool
03:24:55  <thl0>defunctzombie: base64 encode/decode seems to be rather fast (that would be the main bottleneck)
03:25:26  <rvagg>mbalho: what's the word on nodeup? I see no replies to dshaw so far
03:28:02  <defunctzombie>oh man.. I can beat my longest github streak
03:28:04  <defunctzombie>haha
03:28:23  <defunctzombie>oh noes.. substack has me beat!
03:28:31  <defunctzombie>I must foil his streak
03:29:27  <substack>my public+private streak is 51 days
03:29:48  <substack>I actually just passed my earlier stream of 30 days today
03:29:52  <substack>*streak
03:30:05  <defunctzombie>damn you
03:30:26  <defunctzombie>I will have my revenge!!!
03:31:46  <thl0>substack: PR done - once you merge it all the CoffeeScript people will come and use browserify ;)
03:32:13  <defunctzombie>it brings all the noobs to the yard. and their like... its better than ours
03:32:38  <defunctzombie>*they're
03:32:39  <thl0>defunctzombie: :)
03:32:49  <substack>thl0: yep
03:33:45  <thl0>btw substack should I make a PR to make transform() return the browserify instance to allow chaining?
03:35:09  <substack>if you want
03:35:34  <substack>I'm not much a fan of chaining except for short expressions and array pipelines
03:35:40  <substack>but whatevs
03:36:14  <thl0>substack: ok - doesn't hurt anything and is more consistent since require does return the instance
03:36:20  <Raynos>rvagg, dominictarr, hij1nx: whats happening with leveldb/nodeup tomorrow?
03:37:03  <Raynos>personally I would say leave me out of it and have rvagg/dtarr/hij1nx/maxogden do the show
03:37:06  <rvagg>Raynos: I'm going to bring a guitar, you bring some drums, hij1nx can bring a bass and dominictarr can sing, it'll be awesome
03:37:34  <Raynos>i dont know enough about leveldb
03:38:46  <substack>thl0: should I bump the minor on this?
03:38:54  <substack>probably since it's a new feature
03:39:24  <thl0>substack: yep, since it may affect how long bundling takes also and how long people's source maps get
03:41:10  <substack>published
03:41:24  <substack>browserify 2.7.0, browser-pack 0.5.0
03:41:56  <substack>so this is 3 articles I need to write for browserify.org now
03:42:32  <substack>1) transforms 2) source maps 3) static assets
03:43:10  <defunctzombie>thl0: did bundling time increase?
03:43:18  <defunctzombie>thl0: if source maps is off/on?
03:43:22  <thl0>defunctzombie: not noticably
03:43:34  <defunctzombie>what about if they are off?
03:43:35  <thl0>defunctzombie: didn't run any perf tests though
03:43:52  <thl0>defunctzombie: if they are off it shouldn't affect it at all
03:43:54  <defunctzombie>k
03:43:57  <substack>the browserify test suite didn't seem to run noticably slower
03:44:04  <defunctzombie>that is good
03:44:11  <defunctzombie>the jquery perf test is a decent one
03:44:15  <defunctzombie>just bundle the jquery file
03:44:24  <defunctzombie>it is a rather large single file
03:44:25  <thl0>substack: well you'd only see big changes once you source map a huge file
03:44:44  <defunctzombie>d3 is a large project too
03:44:51  <defunctzombie>could try that as well
03:44:52  <thl0>substack: in the sourcemap tests I use small snippets
03:45:15  <chrisdickinson>it'd be neat to have a readable stream that wrote an infinite stream of varied, valid javascript
03:45:18  <thl0>defunctzombie: actually the webgl sample had jquery and three and it bundled in about 5s
03:45:45  <defunctzombie>with sourcemaps?
03:45:53  <defunctzombie>man.. we gotta speed this stuff up somehow :/
03:45:58  <defunctzombie>what is it without?
03:46:10  <thl0>defunctzombie: let me see
03:46:26  <defunctzombie>5 seconds is pretty painful for iterative dev
03:46:35  <defunctzombie>that is why enchilada has a watch mode now
03:46:49  <defunctzombie>some bundles are just too large to only fetch on refresh during dev
03:46:56  <defunctzombie>and also helps with live reload during dev too
03:47:53  <thl0>defunctzombie: without: coffee build 3.71s user 0.15s system 99% cpu 3.879 total
03:48:12  <defunctzombie>oh this uses coffeescript too I forgot
03:48:17  <thl0>defunctzombie: with: coffee build 4.16s user 0.14s system 100% cpu 4.272 total
03:48:37  <defunctzombie>how slow is coffeescript .. wait.. I just realized I don't care hahaha
03:49:08  <thl0>defunctzombie: so not much, but for the coffeescript files we transform with coffeeify which adds sourcemap for each of those no matter if browserify is in debug mode
03:49:26  <thl0>defunctzombie: yeah, this includes coffee script compilation
03:49:33  <thl0>the time I mean
03:50:04  <defunctzombie>yea
03:54:40  * gwenbelljoined
03:56:51  * mikolalysenkojoined
03:57:02  * stlsaintjoined
03:59:28  * st_lukequit (Remote host closed the connection)
04:01:29  <thl0>substack: browserify build broke - found the problem in combine-sourcemap and patched it
04:01:32  * jibayquit (Remote host closed the connection)
04:01:48  * mikolalysenkoquit (Ping timeout: 245 seconds)
04:02:05  <thl0>substack: maybe kick it off again - not sure why I got an arg there that had no replace method
04:04:04  <thl0>substack: or just merge the PR I sent which will kick it off ;)
04:07:47  <substack>ok
04:08:06  <substack>test hook sent
04:08:56  <defunctzombie>thl0: lets make sure we aren't leaking any globals this time haha
04:09:28  * gwenbellquit
04:09:41  <thl0>defunctzombie: you talking about that wrapped thing?
04:09:48  <defunctzombie>:p
04:10:14  <thl0>defunctzombie: yep those things happen since I don't want to put 'use strict' on top of the file and piss you guys off ;)
04:10:30  <defunctzombie>yes, we won't have any of that
04:11:09  <thl0>defunctzombie: which is another good reason to create separate modules, I get to do what I want and can have my own coding style (tw: 2, commafirst)
04:11:25  <defunctzombie>death to commafirst!
04:12:35  <thl0>substack: btw sorry that broke the build, I have no clue why I get something that is not a string there
04:13:50  * marcello3djoined
04:15:29  <substack>I like how having separate modules distributes the job of taking patches
04:15:51  <substack>and how you can make changes in modules without worrying if you'll break other parts since you can just semver
04:20:31  <thl0>substack: ditto that - I guess one lesson learnt here though is that you gotta put it ALL together and run tests before PRing
04:20:56  <thl0>substack: I ran tests on all the modules themselves (including browser-pack), but not with browserify
04:28:55  * thl0quit (Remote host closed the connection)
04:29:25  <defunctzombie>substack: only if minor is properly bumped and not patch
04:29:37  <defunctzombie>people will bump patch for many things tho
04:29:59  <defunctzombie>once the initial patch flow quiets down tho I think it is deff easier to update just one module
04:30:25  * thl0joined
04:31:59  * thl0quit (Remote host closed the connection)
04:35:00  <substack>https://github.com/leapmotion/leapjs/blob/master/Makefile#L25
04:44:12  <defunctzombie>fuck yea thats right bitches
04:44:15  <defunctzombie>use that browserify
04:44:33  <defunctzombie>too bad they don't know how to make makefiles
04:44:37  <defunctzombie>with proper target deps
04:45:10  <defunctzombie>I pre-ordered a leapmotion
04:45:14  <defunctzombie>but am trying to cancel the order
04:45:19  <defunctzombie>cause I don't have time to play with it
04:46:05  <defunctzombie>https://github.com/leapmotion/leapjs/blob/master/package.json#L15
04:46:10  <defunctzombie>they are gonna get fucked by this
04:46:16  <defunctzombie>https://github.com/leapmotion/leapjs/blob/master/package.json#L11
04:46:17  <defunctzombie>and that
04:59:25  <substack>add line-comments to their commit
04:59:47  <rvagg>sounds like you guys should send a friendly PR then
05:07:34  <Raynos>rvagg: thanks :)
05:07:43  * mikolalysenkojoined
05:09:58  <jesusabdullah>dammit Raynos
05:10:06  <Raynos>?
05:10:12  <jesusabdullah>why'd you bail on nodeup -_-;
05:10:45  <Raynos>because hij1nx knows way more about leveldb then I do
05:10:51  <Raynos>thus the session will be better / more interesting
05:11:13  <rvagg>yeah, but nobody has stronger opinions than you so it won't be as entertaining
05:11:25  <jesusabdullah>twitter indicates that you bailed and he stepped in as a replacement
05:11:36  * mikolalysenkoquit (Read error: Operation timed out)
05:12:09  <rvagg>twitter doesn't allow for the subtlety, particularly when many parties are involved in the conversation
05:13:08  <jesusabdullah>>_<
05:13:24  <jesusabdullah>let it be known I would have rather heard Raynos.
05:14:06  <Raynos>rvagg: true dat. But I would have just spouted opinionated bullshit :P
05:14:17  <rvagg>heh, that's 1/2 the fun!
05:14:38  <rvagg>really need dominic on this, hope he shows up today
05:16:42  <Raynos>if he doesnt ill be on standby
05:16:47  <jesusabdullah>ohey while I'm talking to people
05:16:51  <Raynos>or you 3 just do it
05:16:56  <jesusabdullah>who likes literate programming and numerical methods?? https://gist.github.com/jesusabdullah/5178955
05:16:59  <jesusabdullah><--this guy
05:17:36  <substack>having tried literate programming, I like externalized docs more
05:18:06  <jesusabdullah>These are ***not*** docs
05:18:15  <jesusabdullah>If you think of LP as documentation you are doing it wrong
05:18:45  <jesusabdullah>It makes sense in this case because I'm deriving a method and doing an inline translation between theory and code
05:19:34  <jesusabdullah>Being able to pdf-ize this helps with debugging believe it or not
05:19:47  <jesusabdullah>because I can follow the line of reasoning and find out where I done fucked up
05:20:15  <substack>is there a slice stream module?
05:20:23  <jesusabdullah>When it comes to docs I completely agree, inline docs are worthless
05:20:41  <substack>there's a slice-stream on npm but it's not what I had in mind
05:21:03  <substack>I'm more thinking a filter stream that acts like [].slice()
05:21:16  <jesusabdullah>I don't understand?
05:21:31  <jesusabdullah>maybe best to write it
05:21:33  <substack>probs
05:21:46  <substack>a.pipe(slice(3)).pipe(b)
05:21:59  <jesusabdullah>oh, drop the first 3 events?
05:22:00  <substack>would drop the first 3 entries and pipe the rest to b
05:22:03  <substack>yep
05:22:04  <jesusabdullah>sure, I see
05:22:12  <jesusabdullah>slice doesn't sound like the right semantic to me
05:22:18  <jesusabdullah>I'm not one for streams though
05:22:28  <jesusabdullah>I mean I appreciate what they do but I don't geek over them
05:22:50  <jesusabdullah>dammit where is my beer
05:24:26  <jesusabdullah>weirdest new twatter followers ever
05:24:29  <jesusabdullah>some software dude (normal)
05:24:38  <jesusabdullah>a random teenage girl that obviously does not code
05:24:45  <substack>possibly a bot
05:24:46  <jesusabdullah>a parody account of a certain programmer (funny)
05:25:02  <jesusabdullah>a company that sells ??!?!
05:25:07  <jesusabdullah>possibly but the tweets look legit
05:25:10  <jesusabdullah>I checked
05:25:13  <jesusabdullah>I always check
05:25:38  <jesusabdullah>dammit where's st_luke
05:25:41  * tilgoviquit (Remote host closed the connection)
05:25:44  <jesusabdullah>dammit where's my money
05:26:01  <jesusabdullah>dammit why don't I own a house
05:26:20  <jesusabdullah>dammit I'm getting more booze this will not stand!!
05:28:21  <substack>why own a house?
05:31:07  <defunctzombie>substack: sounds like sieve to me
05:31:18  <defunctzombie>I need to book a flight to bay area
05:31:30  <defunctzombie>May 13th is looking like the date
05:33:26  <substack>need a place to crash?
05:34:31  <defunctzombie>I will
05:34:38  <defunctzombie>I want to bounce around :)
05:34:43  <defunctzombie>visit you
05:34:53  <defunctzombie>Raynos, other peeps I know in the area
05:34:57  <defunctzombie>I have a friend at standford
05:36:27  <defunctzombie>I also need to figure out where napa is in relation to san fran
05:36:34  <defunctzombie>I have to attend a wedding june 8th there
05:37:26  <defunctzombie>renting a car is probably gonna have to happen
05:38:32  <substack>the bart runs all the way up to pittsburgh bay point
05:38:39  <substack>can probably take a short bus ride to napa
05:40:44  <jesusabdullah>substack I saw WALNUT CRICK on THE MENTALIST
05:40:58  <jesusabdullah>and I was like I TTLY NOE WHERE THATS AT KIND OF
05:41:31  <defunctzombie>wonder how well I can get around on public transit
05:42:21  * defunctzombiechanged nick to defunctzombie_zz
05:44:10  <jesusabdullah>defunctzombie_zz: fairly well. BART only sucks a little bit and has wide coverage. MUNI really sucks but has comprehensive SF coverage.
05:46:35  * blobaumjoined
05:48:34  <substack>so what is the thing that does rss these days
05:48:37  <substack>I don't even know
05:48:49  <substack>there are so many specs online I don't know what anything means anymore
05:49:29  <Raynos>defunctzombie: let know dates!
05:49:31  <mbalho>how do i take "0x2ecc71" and turn it into 0x2ecc71 in js programatically
05:49:52  <mbalho>parseInt("0x2ecc71", 16)
05:49:52  <mbalho>3066993
05:50:01  <substack>>> Number('0x2ecc71')
05:50:02  <purr>substack: (number) 3066993
05:50:29  <mbalho>i want to do string -> hex literal
05:50:38  <substack>?
05:50:41  <substack>hex literal?
05:50:44  <mbalho>yea type parseInt("0x2ecc71", 16)
05:50:46  <mbalho>oops
05:50:54  <mbalho>type 0x2ecc71 into js repl
05:50:56  <substack>that doesn't make sense to me
05:51:05  <mbalho>oh it converts it to int
05:51:05  <mbalho>HMM
05:51:18  <substack>what are you trying to do?
05:51:31  <rvagg>>> Number("0x2ecc71").toString(16)
05:51:31  <purr>rvagg: (string) '2ecc71'
05:51:32  <mbalho>turn rgb colors into hex colors for three.js
05:51:42  <mbalho>rvagg: oo
05:51:44  <substack>why not just keep them rgb?
05:51:50  * konmjoined
05:51:55  <mbalho>cause webgl is dumb i guess
05:51:55  <substack>rgb(255,255,255) is valid css
05:52:24  <substack>>> '#' + [ 255, 255, 255 ].map(function (c) { return c.toString(16) }).join('')
05:52:24  <purr>substack: (string) '#ffffff'
05:52:29  <mbalho>heh
05:52:32  <substack>except needs padding
05:52:33  <substack>so
05:53:31  <substack>>> '#' + [ 1, 1, 1 ].map(function (c) { return ('0'+c.toString(16)).slice(-2) }).join('')
05:53:31  <purr>substack: (string) '#010101'
05:53:36  <substack>there we go ^_^
05:53:49  <rvagg>btw, there's a growing list of colour libs on microjs: http://microjs.com/#color
05:53:50  <substack>>> '#' + [ 1, 44, 1 ].map(function (c) { return ('0'+c.toString(16)).slice(-2) }).join('')
05:53:50  <purr>substack: (string) '#012c01'
05:54:03  <substack>rvagg: but are they on npm?
05:54:21  <rvagg>substack: some
05:56:33  * dominictarrjoined
05:56:55  <Raynos>rvagg: do you maintain microJS?
05:57:07  <mbalho>i like heathers, npm install color
05:57:38  <rvagg>Raynos: sort of
05:57:55  <Raynos>substack, maxogden, rvagg: we need to write a microJS for browserify compatible modules
05:58:12  * mikealquit (Quit: Leaving.)
05:58:20  <rvagg>Raynos: what you need to do is tag all browserify modules with 'browserify' and they'd show up with microjs.com/#browserify
05:58:32  <rvagg>I've been meaning to do the same for #node and #ender but never get around to it
05:58:59  <Raynos>rvagg: but that isn't good enough. I want microJS style list curation but only contain browserify / npm compatible modules
05:59:10  <rvagg>yeah, fair enough
05:59:11  <Raynos>Actually even further it should favor testling modules
05:59:16  <Raynos>modules that have testling tests
05:59:17  <dominictarr>substack: does your hash-join module work with streams?
06:00:23  <Raynos>Blargh, curated lists
06:00:27  <Raynos>I need to build them :(
06:02:22  * mikealjoined
06:03:55  <Raynos>+
06:16:05  * konmquit (Quit: Colloquy for iPad - http://colloquy.mobi)
06:16:59  <substack>dominictarr: it just has .insert()
06:18:03  <substack>Raynos: I nearly have all the pieces to build that from the browserify.org search
06:18:28  <Raynos>we need a pretty list that's manually curated / categorized
06:18:51  <jesusabdullah>curation sounds like effort
06:18:51  <Raynos>at least until npm search 2.0 and keywords organically start working
06:19:03  <Raynos>its massive effort but worth it for community building
06:20:11  <jesusabdullah>so
06:20:17  <jesusabdullah>I applied for a regular engineering job
06:20:22  <jesusabdullah>with the anchorage power utility
06:20:30  <jesusabdullah>pays like $50/hr
06:20:35  <mbalho>yer in alasker again?
06:20:38  <jesusabdullah>yeah
06:20:40  <jesusabdullah>idk man
06:20:47  <jesusabdullah>I feel like I'm settling if I settle here
06:20:56  <jesusabdullah>like, stifling some sort of potential
06:20:58  <jesusabdullah>but I love it here
06:21:10  <jesusabdullah>I'll be here through the summer anyway (brother's getting married)
06:21:19  <jesusabdullah>but man oh man do I want a permanent house
06:21:28  <jesusabdullah>and a well-paying gig
06:21:37  <jesusabdullah>and on some level I'm kind of getting tired of software
06:21:45  <jesusabdullah>I get burnout pretty easy s
06:21:47  <jesusabdullah>so
06:21:48  <jesusabdullah>idk
06:21:53  <jesusabdullah>I'm a self-doubter
06:21:56  <substack>you could just do ad-hoc consulting gigs
06:21:57  <substack>remotely
06:22:07  <substack>whenever you feel like working
06:22:26  <jesusabdullah>I feel like stability
06:22:30  <jesusabdullah>that's what I feel like
06:22:36  <jesusabdullah>but at WHAT COST substack
06:22:38  <jesusabdullah>AT WHAT COST
06:23:32  <jesusabdullah>AL JAZEERA: KEEPIN' IT REAL SINCE 1996
06:24:58  <jesusabdullah>really I wish I knew what I wanted
06:26:28  <jesusabdullah>fewer whims
06:26:49  <jesusabdullah>what about you guys? Do you ever get in a, "what the fuck am I doing with my life" mood?
06:26:53  <jesusabdullah>How do you stave that shit off?
06:27:59  <substack>go do something silly
06:28:47  <jesusabdullah>I already did that, remember?
06:28:53  <jesusabdullah>I've been doing something silly since 2011
06:29:11  <substack>https://en.wikipedia.org/wiki/Meaning_(existential)
06:29:26  <substack>also https://en.wikipedia.org/wiki/Absurdism
06:29:28  <jesusabdullah>sometimes I wonder if I was in a sane state of mind when I moved
06:29:32  <dominictarr>jesusabdullah: software is most fun when you are just doing it for fun
06:29:38  <jesusabdullah>did you know I have bipolar disorder substack?
06:29:42  <substack>In philosophy, "the Absurd" refers to the conflict between the human tendency to seek inherent value and meaning in life and the human inability to find any.
06:30:12  <jesusabdullah>ugh fuck I must be drunk if I'm being that open in a public forum >_<
06:30:35  <jesusabdullah>substack: I think Marak is an absurdist, he's tried to explain that to me a few times
06:30:42  <substack>"Absurdism, therefore, is a philosophical school of thought stating that the efforts of humanity to find inherent meaning will ultimately fail (and hence are absurd) because the sheer amount of information as well as the vast realm of the unknown make certainty impossible."
06:31:14  <jesusabdullah>I mean, I don't really believe in a meaning of life. But, I do believe in some optimum set of choices which will make me the "most happy"
06:31:21  <jesusabdullah>or the "most content"
06:31:41  <jesusabdullah>dominictarr: that's occurred to me. If I take an engineering job there *is* that
06:31:53  <Raynos>jesusabdullah: i work on open source for fun
06:31:59  <jesusabdullah>oh dominictarr did I show you my literate programming thing y/n
06:32:02  <Raynos>and when I dont work on open source because im burnt out i just rot
06:32:08  <dominictarr>jesusabdullah: no, link?
06:32:11  <Raynos>I just know how to not take depression seriously
06:32:23  <jesusabdullah>dominictarr: https://gist.github.com/jesusabdullah/5178955 noweb + js + numerical methods
06:32:42  <jesusabdullah>= a small chunk of dc-tycoon's eventual customers AI
06:32:48  <substack>jesusabdullah: you could put together a book
06:32:55  <dominictarr>jesusabdullah: how do I run it?
06:33:06  <substack>with silly cartoons and some topic of interest
06:33:45  <substack>make a kickstarter for it like http://www.kickstarter.com/projects/creationix/js-git
06:33:46  <jesusabdullah>dominictarr: https://github.com/jesusabdullah/data-center-tycoon/blob/master/Makefile shows how to convert it to javascript that's require-able
06:34:11  <jesusabdullah>dominictarr: https://github.com/jesusabdullah/data-center-tycoon/blob/master/test.js this test program uses a module I was toying with which does more or less the same thing
06:34:20  <dominictarr>jesusabdullah: can you make a github pages site for it?
06:34:50  <jesusabdullah>dominictarr: You mean, make the pdf output a github page?
06:35:06  <substack>jesusabdullah: you could make a web game about engineering
06:35:14  <dominictarr>oh, it's a pdf output?
06:35:18  <substack>building bridges and such!
06:35:27  <jesusabdullah>dominictarr: yeah, the weave is LaTeX
06:35:39  <substack>with real stress forces and such
06:35:41  <Raynos>jesusabdullah: why does that look like latex?
06:35:44  <jesusabdullah>dominictarr: noweb can do html instead of tex but in this case I needed to typeset equations
06:35:53  <Raynos>substack: also wow, creatonix actually got it funded, epic
06:35:54  <jesusabdullah>Raynos: cause it's LaTeX with javascript embedded in
06:36:04  <substack>Raynos: in less than 24 hours!
06:36:21  <Raynos>man
06:36:28  <jesusabdullah>Raynos: running the right commands pulls out the javascript and assembles it properly, running others turns it into a pretty-printed pdf document
06:37:16  <Raynos>aw man
06:37:21  <Raynos>the idea of coding in latex is crazy
06:37:25  <Raynos>your code doubles as a pdf
06:37:39  <jesusabdullah>that's old-school literate programming for you
06:37:49  <jesusabdullah>the original conception was by don knuth of LaTeX fame
06:37:58  <jesusabdullah>or rather, TeX fame
06:38:31  <jesusabdullah>in fact, the TeX source code is written in web, Knuth's implementation of a literate programming tool, and could be compiled into either a .dvi (readable) or into pascal source
06:38:35  <substack>knuth of "schooled by McIlroy" fame >:D
06:38:53  <jesusabdullah>https://github.com/jesusabdullah/data-center-tycoon/blob/master/solver.pdf?raw=true
06:39:07  <jesusabdullah>I'm unfamiliar with your reference sub
06:40:21  <hij1nx>dominictarr: fun! levelweb can now build a zoomable tree-map by grouping your keys based on the character you use to divide them -- http://d.pr/i/BZP7
06:40:53  <substack>jesusabdullah: http://www.leancrew.com/all-this/2011/12/more-shell-less-egg/
06:41:00  <dominictarr>hij1nx: OH SWEET
06:41:18  <dominictarr>hij1nx: what character do you use?
06:41:18  <hij1nx>dominictarr: it really gives you an idea of whats in your database
06:41:25  <substack>knuth wrote 10 pages of literate pascal with bugs, mcilroy wrote a bug-free 6-line shell script to do the same thing
06:41:27  <hij1nx>i use a pipe
06:41:32  <hij1nx>dominictarr: ^
06:41:36  <substack>and 6 lines was only because each command got its own line
06:41:55  <hij1nx>dominictarr: but you can specify any character
06:42:00  <jesusabdullah>I believe it substack, tex was smart but pascal kinda sucks
06:42:02  <dominictarr>hij1nx: use ~ or !
06:42:16  <dominictarr>(because thef are the last, and first printable characters)
06:42:33  <hij1nx>dominictarr: for dividing your keys?
06:42:40  <dominictarr>yeah
06:42:44  <jesusabdullah>and LP isn't always the answer, often it obscures more than it clarifies
06:42:52  <hij1nx>dominictarr: im talking about `foo:bar:bazz` as a key name
06:42:58  <hij1nx>dominictarr: where `:` is the divider
06:43:36  <hij1nx>dominictarr: i think you are talking about for searching?
06:44:00  <hij1nx>dominictarr: { start: 'foo:bar', end: 'foo:~' }
06:44:02  <dominictarr>the problem with : is that it is in between numbers and letters, so it will beak ordering when you have numbers in your keys
06:44:36  <dominictarr>I always separate meta stuff away with ~
06:44:48  <dominictarr>so you know that stuff will appear after the main content
06:44:48  <hij1nx>dominictarr: i guess pipe will have the same problem
06:44:54  <dominictarr>yup
06:45:07  <hij1nx>i've been using pipe, but now i will switch to tilde
06:45:37  <hij1nx>i knew this about tilde since i use it as a search terminal
06:45:37  <dominictarr>I think the best is ~ for "sections"
06:45:51  <dominictarr>and then if you have ordered groups
06:45:52  <dominictarr>use !
06:45:54  <hij1nx>dominictarr: problem is that people like ":"
06:45:55  <chilts>I'm currently using / for sections, but ~ for the end of a search
06:45:59  <chilts>maybe I should switch too :)
06:46:07  <dominictarr>foo!thing, foo2!thing_ fo!thing
06:46:09  <hij1nx>chilts: if ordering maters
06:46:12  <dominictarr>that will sort correctly
06:46:14  <chilts>is prior to both numbers and letters I think
06:46:21  <dominictarr>hij1nx: too bad
06:46:36  <dominictarr>I didn't set the ordering of ascii
06:46:49  <hij1nx>dominictarr: not really, its a cultural thing, a lot of people do it in redis
06:47:16  <hij1nx>dominictarr: im not particular on an aesthetic for what character i use
06:47:40  <hij1nx>dominictarr: currently, i've not had any problems with sorting, but this is good to keep in mind
06:47:54  <dominictarr>heh, so ascii is a telegraph thing - I guess they didn't envision leveldb when they designed ascii
06:48:59  <hij1nx>dominictarr: next up im going to pull the timeseries stuff into levelweb, its really nice to be able to see the data that is getting put into the database as it gets put in.
06:49:57  <hij1nx>its a little messy atm, but a good start -- https://github.com/hij1nx/levelweb
06:49:58  <dominictarr>hmm, totally
06:50:24  <hij1nx>gets the basic ideas out of my head.
06:50:33  <chilts>I think in my case, using / is okay, since the keys are similar to 'itemName/timestamp...etc...' rather than 'itemName/attributeName'
06:50:39  <dominictarr>hij1nx: oh, I was messing with this thing yesterday - that turns an ascii key into a hsl color
06:51:02  <dominictarr>so that the sorting approximately maps to the rainbow
06:51:13  <hij1nx>coool!
06:53:10  <dominictarr>it's just a proof of concept right now, but i'll tidy it up soon
06:55:43  * owen1joined
07:06:43  <rvagg>hij1nx: lev needs to be 0.10ified
07:06:56  <rvagg>path.join() is given an array when you specify db location, 0.10 doesn't like that
07:10:40  <chilts>if you're talking Node v0.10, it doesn't work on my v0.8.22 either :/
07:10:46  <chilts>so not sure it's v0.10 specific
07:12:30  <rvagg>mmm, chilts, hij1nx, I'm getting an internal leveldb fail that's killing my node process and I have no idea how it's managing to do that! shouldn't be possible to do but there's obviously a bug somewhere
07:12:53  <rvagg>something to do with iterators/readstream
07:14:20  <rvagg>arrrrr! chilts, hij1nx, it's an encoding issue, lev assumes json by default but I'm only doing utf8 on for both keys and values on this db
07:14:24  <chilts>with lev?
07:14:33  <rvagg>but still, it should be able to crash a node process so there's something to fix for me
07:14:39  <chilts>right
07:14:44  <rvagg>chilts: yes, with lev, if I tell lev just to go with utf8 then it's fine
07:14:52  <chilts>I'm not sure what you mean, but sounds like you're onto it :)
07:15:15  <chilts>ah, I'll try with telling lev to do utf8, even though pretty much all my keys and values are ascii
07:16:03  <rvagg>chilts: lev --encoding utf8
07:23:20  <chilts>same error
07:23:31  <chilts>sorry, my pizza is cooked, back soon :)
07:24:56  <dominictarr>substack: what was your module that matched paths into objects like in JSONStream? I can't find it?
07:31:18  <substack>https://github.com/substack/node-pathway
07:32:11  <substack>mostly I used that module to experiment with the JSONStream parser true/false and function syntax
07:32:30  <substack>they're at parity since you merged that pull request a while back
07:32:57  <dominictarr>thanks
07:33:40  <dominictarr>substack: oh, JSONStream now also has 'rows.*.value' syntax if it's a string
07:34:05  <dominictarr>substack: and, today, a patch that handles very large json better
07:34:19  <dominictarr>parsed a 1gb json
07:35:03  <substack>does that fix memory leaks with the root object?
07:35:16  <substack>because that is exactly the issue that testling-ci was having earlier today
07:35:38  <substack>we have some *very large* streams of json
07:46:50  <rvagg>juliangruber: there's a couple of methodology questions for you @ https://gist.github.com/juliangruber/5139252
07:47:22  <substack>mikeal: you can add me to planet node now http://substack.net/blog.xml
08:01:52  <chrisdickinson>dominictarr: you wouldn't happen to have a link to that very large json handy, would you?
08:02:16  <dominictarr>chrisdickinson: there is a script to generate it in the repo
08:02:21  <dominictarr>test/gen.js
08:02:22  <chrisdickinson>ah, cool
08:02:38  <chrisdickinson>i've got https://github.com/chrisdickinson/json-parse-stream, and it'd be nice to make sure i don't have any leaks :|
08:02:40  <dominictarr>also, a good one is all of npm
08:03:11  <dominictarr>curl isaacs.iriscouch.com/registry/_all_docs?include_docs=true
08:03:43  <dominictarr>this was about 350 mb when I dl'd it a few weeks ago
08:04:13  <substack>blarg there is no rfc3339 date modules on npm
08:04:15  <substack>writing one
08:04:38  <dominictarr>hmm, is github down for you guys?
08:05:14  <substack>it's up
08:05:16  <CoverSlide>https://status.github.com/
08:05:21  <CoverSlide>wfm
08:06:45  <dominictarr>hmm, I can see that page, but I get connection timed out on everything else
08:09:03  <dominictarr>and I can push to git, weird
08:16:42  <dominictarr>chrisdickinson: will be interesting to bench mark the parsers on large json!
08:29:23  * No9quit (Ping timeout: 256 seconds)
08:56:06  <substack>daaaamn timestamp locales are some crazy shit https://ci.testling.com/substack/internet-timestamp
08:56:44  <substack>having an automated test service shows me routinely how fucked up browsers actually are
09:05:49  * No9joined
09:09:52  <dominictarr>hmm, what would it take to put all the cli tools like gcc and git and all that into npm?
09:10:43  <rvagg>need a kickstarter for each to rewrite them into js
09:15:17  * No9quit (Read error: Connection reset by peer)
09:15:27  <dominictarr>well, I think creationix can probably handle git.js
09:15:57  <juliangruber>new little tool: http://ghub.io/
09:16:44  <dominictarr>juliangruber: NICE!
09:17:28  <juliangruber>dominictarr: :)
09:17:38  <dominictarr>if we can get everything into npm, then we can install everything into node_modules instead of the dumb $PATH shit
09:18:55  <juliangruber>dominictarr: like phantomjs already being available through npm
09:29:58  <chrisdickinson>dominictarr: i've almost got the reading side of git.js done
09:30:17  <chrisdickinson>though zlib-browserify doesn't implement inflate ):
09:30:39  <dominictarr>does it work on node?
09:31:17  <chrisdickinson>yeah. i'm reimplementing my old node git library (https://github.com/chrisdickinson/tempisfugit) as a bunch of tiny libraries
09:37:14  <substack>instead of paginating a stream of results from couchdb
09:37:34  <substack>maybe I can use backpressure?
09:37:48  <substack>not sure how this would work communicating backpressure from browser code
09:38:06  <substack>I would need to serialize the pause and resume signals explicitly over the wire
09:40:46  <dominictarr>substack: mux-demux does that already!
09:42:06  <dominictarr>it sends pause and resume messages --- but you could also have another system that worked like tcp - with naks and acks, and that adjusted speed - that could also be quite good for degrading notifications that arn't so urgent
09:42:31  <substack>oh most excellent
09:43:05  <substack>well I'll need to have a custom thing to paginate couch still to send startkey/limit values accordingly
09:43:14  <substack>but at least the external interface can just be a stream
09:43:20  <substack>that simplifes the UI code a ton
09:45:36  <dominictarr>I had a pagination stream on my todo list once, thought I could inject closure to customize it to work with anything with pages
09:46:14  <dominictarr>I think it dropped off my list for some reason, I think because it didn't turn out to be as simple as I thought
09:46:25  <dominictarr>if you just do couch, no problem though.
09:47:58  <substack>I just need it so our servers don't melt
09:48:13  <substack>earlier today testling-ci was down because a repo had too much data
09:48:32  <substack>and the algorithms I wrote to parse couch were just taking in all the data
09:55:39  <dominictarr>substack: rewrite with leveldb!
09:56:08  <substack>might be worth it
09:56:30  <substack>is rumours in a good enough state to start using as a json bucket?
09:57:14  <substack>this could save me some trouble writing the realtime update feed
10:00:43  <substack>it seems that the basic approach to datatypes fits really well with what I need too
10:04:04  <dominictarr>hmm, sure - I'd probably use crdt for the test state - which would show pending, etc really nice
10:04:33  <dominictarr>and then keep log output (which might get way larger) as ranges in leveldb
10:04:49  <dominictarr>TODO: write module for storing streams in leveldb.
10:14:29  <juliangruber>dominictarr: hij1nx and me need this for tsd too, just a bit more powerful: store a stream of numeric values as time series data in leveldb. Then: range queries and getting data in defined intervals (we don't need 1s data when displaying data for 1year)
10:14:46  <juliangruber>so the 2nd part can build on top of your module
10:14:58  <juliangruber>yours just had to assign timestamps to entries
10:15:04  <chrisdickinson>woo!
10:15:14  <dominictarr>juliangruber: so that is just a merge/reduce thing
10:15:18  <chrisdickinson>reading out of packfiles works
10:15:48  <juliangruber>dominictarr: I see, yes
10:15:51  <rvagg>juliangruber: yeah, wolfeidau is doing something very similar for storing stats and is using map-reduce for that
10:16:22  <dominictarr>juliangruber: rvagg you could probably optimize this for time series
10:16:23  <juliangruber>rvagg: he should jump on board of hij1nx/tsd then :)
10:16:32  <dominictarr>juliangruber: what is tsd?
10:16:46  <rvagg>juliangruber: yeah I've pointed him that way, not sure how he's going with tsd or whether his work is totally separate
10:17:06  <juliangruber>dominictarr: https://github.com/hij1nx/tsd
10:17:17  <dominictarr>aha, I see
10:17:48  <dominictarr>you could do things like aggregate to different levels of granularity
10:18:04  <dominictarr>but take into account the time
10:18:17  <dominictarr>like, aggregate the last minute only every minute, etc...
10:19:11  <dominictarr>and then prefix each level so that it's you can aggregated data really fast.
10:21:48  <dominictarr>hmm, actually.. map-reduce might already work pretty well for that except that without the right timing
10:22:15  <dominictarr>the current aggregation levels will be in accurate until they are finished
10:29:50  <juliangruber>disk space should be cheaper than processing power
10:29:57  <juliangruber>so the 1st approach can be better
10:29:59  <juliangruber>except
10:30:04  <juliangruber>you can't change granularities afterwards
10:30:16  <juliangruber>or if you change granularities it has to do a lot of work
10:34:41  <dominictarr>juliangruber: recalculating the aggregations isn't a big job
10:34:46  <chrisdickinson>follow up woo: npm.im/git-packfile
10:36:22  <dominictarr>chrisdickinson: the repo link is wrong
10:39:59  <rvagg>chrisdickinson: was it you that was playing with protobuf in js?
11:35:46  * ins0mniajoined
12:24:22  * AvianFlujoined
13:09:17  * thl0joined
13:18:29  * thl0quit (Remote host closed the connection)
13:19:07  * thl0joined
13:23:15  * mikolalysenkojoined
13:23:50  * thl0quit (Ping timeout: 256 seconds)
13:40:36  * mikolalysenkoquit (Ping timeout: 248 seconds)
13:54:20  * thl0joined
13:57:59  * mikolalysenkojoined
14:06:29  * nichola__quit (Read error: Connection reset by peer)
14:07:00  * nicholas_joined
14:28:57  * yorickjoined
14:28:57  * yorickquit (Changing host)
14:28:57  * yorickjoined
14:49:45  * konmjoined
14:56:35  * mikealquit (Quit: Leaving.)
15:07:11  * konmquit (Remote host closed the connection)
15:20:54  <juliangruber>I just realised that a PaaS is basically a devops framework. There shuold be libs/apps that do the same thing in a modular way
15:39:52  * mikolalysenkoquit (Ping timeout: 256 seconds)
15:44:41  <hij1nx>juliangruber: it would be cool if we could figure out a way to make the core of tsd a plugin so that it could be consumed by levelweb
15:45:45  <juliangruber>hij1nx: for now everything it does is store the last value per metric in leveldb, right?
15:46:01  <juliangruber>hij1nx: if we do the full timeseries thing than definitely make it a plugin
15:55:51  * defunctzombie_zzchanged nick to defunctzombie
16:11:53  <defunctzombie>http://www.reddit.com/r/node/comments/1afxr9/application_error_logging/
16:11:55  <defunctzombie>upboats!
16:32:34  * mikolalysenkojoined
16:37:43  <juliangruber>$ node server.js 2> >(chunk-stream -c "mail -s "Error" mail@host.com")
16:37:50  <juliangruber>what do you guys think?
16:38:16  <juliangruber>this would send me an email when something gets written to stderr
16:38:20  <juliangruber>in a reasonable interval
16:40:25  <juliangruber>(I'm about to write chunk-stream or however I can call it)
16:42:20  <defunctzombie>juliangruber: sure, but I still favor the approaches I talked about
16:42:30  <defunctzombie>for errors
16:42:56  <defunctzombie>if I use your method I have to add that to everything I run
16:42:59  <juliangruber>I don't need that much, I just want to get notified when there's something on stderr
16:43:00  <defunctzombie>this is operational overhead
16:43:11  <juliangruber>you have to include the library everywhere too
16:43:18  <defunctzombie>juliangruber: that is much easier
16:43:23  <defunctzombie>since it is alreayd in the code
16:43:37  <defunctzombie>no typing mistake or such
16:44:05  <juliangruber>you wouldn't type this by hand but rather include it into your devops ecosystem
16:44:05  <defunctzombie>ive done ops, ive done deploying, etc and I can tell you that the above is not a good way to deploy things
16:44:14  <defunctzombie>it is too fragile
16:45:11  <defunctzombie>you should write to stderr/out and have another process gather that if you want to do that type of approach is my recommendation
16:45:30  <defunctzombie>anyhow, having the logging lib is easier and gives me more info
16:45:36  <defunctzombie>you can use both approaches
16:45:46  <juliangruber>mhm
16:45:47  <defunctzombie>use the logging lib to log sensible things to stdout in whatever format you want
16:45:54  <defunctzombie>and then read that
16:47:14  * yorick_joined
16:47:44  * yorickquit (Ping timeout: 252 seconds)
16:48:02  <defunctzombie>juliangruber: honestly I just recommend doing whatever you are happy with :)
16:48:11  <defunctzombie>juliangruber: I only say all this stuff as tips
16:48:14  * thl0quit (Remote host closed the connection)
16:48:26  <defunctzombie>which I have found very useful having done many deployments and run various systems
16:48:45  <defunctzombie>many of which have to do a lot but be monitored by a small number of devs/ops
16:49:35  <defunctzombie>so everything I say isn't some ideal do this way cause I said so, it is more of a I do this and it has solved problems X, Y, Z for me
16:49:39  <juliangruber>I don't need sentry
16:49:44  <juliangruber>I don't want to spend mony
16:49:46  <juliangruber>money
16:49:56  <juliangruber>can book send emails?
16:51:12  * mikolalysenkoquit (Ping timeout: 256 seconds)
16:51:45  <juliangruber>see, I have webapps that are opensource. When I deploy them I want to get emails. When others deploy them I don't want to get emails from them. So it already has to be in the env vars or argv. Then this chunk-stream thingy isn't much overhead
16:51:56  <juliangruber>so we propably just have different use cases
16:54:09  * st_lukejoined
16:54:19  <st_luke>http://i.imgur.com/cMOwnkW.gif
16:54:26  <defunctzombie>juliangruber: sentry has a free version
16:54:30  <defunctzombie>juliangruber: book can do anything
16:54:48  <defunctzombie>juliangruber: https://github.com/shtylman/node-book-email
16:55:01  <defunctzombie>I don't currently use that module with my book setup, but that is easy enough to do
16:55:12  <defunctzombie>log.ues just takes any function you want to call
16:55:24  <defunctzombie>st_luke: I cant' stop watching that
16:55:45  <defunctzombie>juliangruber: you don't have to get their emails
16:55:53  <defunctzombie>juliangruber: just don't put your email creds in the source
16:56:03  <defunctzombie>juliangruber: I used ENV vars for sentry or such things to activate them
16:56:13  <defunctzombie>so I don't share my credentials to those things
16:56:22  <defunctzombie>and it degrades nicely for those running it locally too
16:56:48  <defunctzombie>do what works for you :)
16:57:40  * mikolalysenkojoined
17:11:31  * defunctzombiechanged nick to defunctzombie_zz
17:28:47  * No9joined
17:31:32  * jibayjoined
17:35:39  * AvianFluquit (Remote host closed the connection)
17:39:07  <mbalho>dominictarr: are you awake?
17:40:35  * shamajoined
17:41:14  * marcello3dquit (Remote host closed the connection)
17:44:53  * shamaquit (Remote host closed the connection)
17:47:22  * thl0joined
17:53:26  <rvagg>mbalho: are we on for 11 or 1? I see @ffloat is cool with 1
17:54:18  <mbalho>rvagg: yep
17:54:23  <rvagg>ahh, dshaw just tweetered, 1
17:54:32  <rvagg>it's nearly 11 now right?
17:54:47  <mbalho>yea 10:55
17:54:55  <rvagg>ok, cool, I'm going back to bed then
17:55:00  <mbalho>heh
17:57:55  <st_luke>is this the most geographically dispersed nodeup ever?
17:58:02  <st_luke>lots of time zones
18:00:06  * mikealjoined
18:01:17  * mikealquit (Client Quit)
18:02:01  * mikealjoined
18:02:19  * mikealquit (Client Quit)
18:02:28  * mikealjoined
18:06:50  * mikealquit (Ping timeout: 256 seconds)
18:17:46  * defunctzombie_zzchanged nick to defunctzombie
18:17:59  * thl0quit (Remote host closed the connection)
18:23:17  <defunctzombie>juliangruber: you know what could be cool, a ghub.io/docserv merging of sorts
18:24:41  <juliangruber>defunctzombie: docserv automatically installs packages it doesn't already have and display docs?
18:24:54  <defunctzombie>juliangruber: nope
18:25:05  <defunctzombie>right now docserv only shows docs for what is installed
18:25:10  <defunctzombie>and lets you know hwat is missing
18:25:11  <defunctzombie>or whatnot
18:25:30  <juliangruber>I meant, that's what it would need to do if I get you right
18:25:34  <defunctzombie>ripping the relevant parts out of tryme
18:25:42  <defunctzombie>to do the install
18:25:47  <defunctzombie>and then running docserv
18:25:50  <defunctzombie>on top of that
18:25:52  <defunctzombie>hahaha
18:26:06  <juliangruber>yeah, would be nice
18:27:22  <defunctzombie>and then we can make npm docs got to that
18:27:27  <defunctzombie>and it would be super win!
18:27:31  <defunctzombie>*go to
18:29:34  * mikolalysenkoquit (Ping timeout: 245 seconds)
18:29:49  <juliangruber>:O
18:29:55  <juliangruber>there is package.repository
18:30:12  <juliangruber>so just get the package from the npm api and display the readme with gfm or something like that
18:30:22  <juliangruber>no need to fully install the package
18:31:06  * thl0joined
18:31:31  * st_lukequit (Remote host closed the connection)
18:36:59  <defunctzombie>juliangruber: yea, that could work too, npm install is nice cause it will also handle github url and such
18:38:05  <juliangruber>npm.get(/<pkg.name>, function (err, pkg) { console.log(pkg.readme) }) should work
18:51:51  * marcello3djoined
18:56:11  * marcello3dquit (Ping timeout: 260 seconds)
18:56:12  * AvianFlujoined
19:01:04  <defunctzombie>pkrumins: https://github.com/shtylman/away new https stuff is great!
19:01:20  <defunctzombie>when the badges have pending browsers on them it will be even better!
19:01:37  <defunctzombie>oh man.. just had a crazy idea!
19:01:47  <substack>working on that part
19:01:48  <pkrumins>glad you're liking https badges
19:01:50  <defunctzombie>substack: gif badges so you can push the changes
19:02:06  <defunctzombie>the browser will keep trying to load the gif as it updates
19:02:24  <substack>are there any good modules for doing this?
19:02:29  <defunctzombie>I don't think so
19:02:33  <defunctzombie>but I haven't checked
19:02:37  <pkrumins>node gif
19:02:42  <pkrumins>but it works only in node 0.2.x
19:02:49  <defunctzombie>and browsers will keep trying to load the gif as long as it hasn't finished iirc
19:02:55  <defunctzombie>but it may need the size upfront, not sure
19:02:57  <pkrumins>we can run node 0.2.x for that
19:02:57  <defunctzombie>wow
19:03:13  <substack>I looked into the framing protocol for gif
19:03:35  <substack>it's not too complicated, just an lzw pass on the data chunks and some stuff in the header
19:04:20  <defunctzombie>I am looking at the wiki page
19:04:27  <pkrumins>nah we can just use node gif with node 0.2.x
19:04:31  <defunctzombie>doesn't seem that the image size needs to be known upfront
19:04:51  <defunctzombie>in either case, if you do need to know the number of frame
19:04:54  <defunctzombie>you know that
19:04:57  <pkrumins>https://github.com/pkrumins/node-gif#animatedgif
19:05:01  <defunctzombie>since the number of browsers is deterministic
19:05:37  <defunctzombie>pkrumins: probably easy enough to update for node 0.8/0.10
19:06:00  * mikolalysenkojoined
19:06:33  * marcello3djoined
19:06:41  <defunctzombie>this sounds like it could work
19:06:42  <defunctzombie>haha
19:06:49  <defunctzombie>the best badge ever
19:13:06  <dominictarr>mbalho: Raynos so, did they shift the nodeup time, I just woke up?
19:14:13  <pkrumins>defunctzombie: very hard to update
19:14:26  <defunctzombie>pkrumins: why is that?
19:14:48  <dominictarr>hmm, no reply. I'm going back ti bed
19:14:50  * dominictarrquit (Quit: dominictarr)
19:14:56  <pkrumins>too many things have changed
19:15:11  <pkrumins>it took me a long time to get just the synchronous part of node-png working in node 0.8
19:15:23  <pkrumins>i've no idea how to do the asynchronous part
19:15:24  <defunctzombie>hm
19:15:29  <defunctzombie>oh, I do haha
19:15:36  <defunctzombie>the apis changed
19:15:41  <mbalho>where did dominic go..
19:15:43  <defunctzombie>but the general ideas didn't
19:17:05  <defunctzombie>if the gif usecase is simple tho, it may not be worth updating and just making a smaller module that does just what is needed for an animated gif in js land
19:17:15  <defunctzombie>really depends on how complex it is to make a gif
19:17:19  <pkrumins>no it's writtten in c++
19:17:37  <pkrumins>uses giflib i think
19:17:44  <defunctzombie>yea, your module does
19:17:49  <mbalho>gifsicle ftw
19:17:58  <defunctzombie>I am saying in general tho for making a gif
19:18:17  <defunctzombie>giflib seems to be pretty small
19:18:22  <defunctzombie>(if I found the right source)
19:20:22  <defunctzombie>wow, not a single other gif npm module out there
19:20:23  <defunctzombie>heh
19:20:29  <defunctzombie>as if people don't work with gifs at all
19:21:02  <defunctzombie>anyhow, substack if you want to do the gif browser badge, I can look at updating the c++ stuff
19:21:32  <substack>sure thing!
19:21:40  <substack>I'm going to get the pending part working today
19:22:13  <substack>if that module gets updated then I'll make a /user/repo.gif route for animated updates
19:22:18  <defunctzombie>cool
19:22:43  <defunctzombie>I am thinking I will make a giflib module which will be the c lib bindings (as minimal as possible)
19:22:48  <defunctzombie>and then update the gif module to use that
19:23:03  <defunctzombie>so it can just be a higher level abstraction or whatnot
19:23:36  <defunctzombie>I maintain the libxmljs bindings so much of this stuff is relatively familiar to me (will have to relearn some things) haha
19:26:42  * nk109joined
19:27:30  * nk109quit (Client Quit)
19:28:11  <pkrumins>defunctzombie: it actually looks like someone patched node-gif to use eio https://github.com/pkrumins/node-gif/pull/4
19:30:42  * ins0mniaquit (Ping timeout: 256 seconds)
19:33:56  * mikolalysenkoquit (Ping timeout: 255 seconds)
19:34:49  <juliangruber>defunctzombie: I wanted to do that for https://github.com/juliangruber/traffic-light too
19:34:59  <juliangruber>it's a shame that we have no gifsockets implementation in node
19:38:08  * No9quit (Quit: Gone to join the choir invisibule)
19:39:16  * mikolalysenkojoined
19:43:44  * marcello3dquit (Remote host closed the connection)
19:44:36  * tim_smart|awayquit (Ping timeout: 245 seconds)
19:45:42  <defunctzombie>pkrumins: yea, but that is now old api hahaha
19:45:49  <defunctzombie>pkrumins: it all needs to use libuv now
19:47:10  <pkrumins>ah
19:58:34  * defunctzombiechanged nick to defunctzombie_zz
20:00:31  * ins0mniajoined
20:05:09  * stlsaintquit (Ping timeout: 256 seconds)
20:07:01  <Raynos>dominictarr: not sure
20:16:49  * marcello3djoined
20:18:21  <Raynos>mbalho: many robots voice
20:19:16  <chrisdickinson>thl0: you in portland?
20:19:35  <thl0>chrisdickinson: nope - NYC
20:19:39  <chrisdickinson>ah, cool
20:20:03  <thl0>chrisdickinson: why, what did I miss ?
20:20:11  * AvianFluquit (Remote host closed the connection)
20:20:13  <chrisdickinson>ah, nothing! just saw your nodepdx submission
20:20:19  <chrisdickinson>the local node group is called pdxnode
20:20:25  <chrisdickinson>which… was confusing
20:20:33  * mmckeggjoined
20:20:39  <thl0>chrisdickinson: ah, hope it gets accepted - this talk was well received here at NYCJS
20:20:43  <chrisdickinson>awesome
20:20:53  * chrisdickinsonsheepishly deletes his PR comment
20:21:43  <chrisdickinson>slides look great
20:21:52  <thl0>chrisdickinson: not fair - didn't even see the comments yet
20:21:54  <chrisdickinson>curious about the one where "use strict" is added in a code example
20:21:54  <thl0>chrisdickinson: thanks
20:21:59  <chrisdickinson>i just +1'd it :)
20:22:03  <chrisdickinson>which is what we do for pdxnode
20:22:18  <thl0>chrisdickinson: plus1 it anyways ;)
20:22:36  <chrisdickinson>haha :D
20:23:07  <thl0>chrisdickinson: trying to teach people the right thing - kinda like if all cool skateboarders wear no helmet, the kids won't either ;)
20:23:08  <chrisdickinson>was thinking about submitting a talk about streams in the browser, even though that's been done before
20:23:33  <thl0>chrisdickinson: you should, sounds interesting
20:23:34  <chrisdickinson>haha
20:23:34  <chrisdickinson>yeah
20:23:57  * juliangruberquit (Read error: Connection reset by peer)
20:24:13  <chrisdickinson>did a bit about it for a pdxnode talk
20:24:21  <chrisdickinson>https://github.com/PDXNode/pdxnode/blob/master/2013-mar/modules/part-3/main-0.js -> https://github.com/PDXNode/pdxnode/blob/master/2013-mar/modules/part-3/main-1.js
20:24:28  <chrisdickinson>5 line autocomplete widget
20:24:39  * marcello3dquit (Remote host closed the connection)
20:24:40  * juliangruberjoined
20:26:04  <thl0>chrisdickinson: cool, all this reminds me of what we did with RX in WPF (C#)
20:26:42  <chrisdickinson>yeah, i really need to explore what's possible with gozala/reducers (which seems like a more "node"-y version of rxjs)
20:26:44  <thl0>chrisdickinson: we would listen on a data stream - possibly combine latest, etc. and the update viewmodel props to show UI changes and vice versa
20:26:56  <chrisdickinson>awesome :D
20:28:33  <thl0>chrisdickinson: btw tried replpad yet? http://thlorenz.github.com/replpad/
20:28:44  <thl0>used as an example in the talk
20:29:03  <chrisdickinson>no, but looks very interesting
20:29:07  <thl0>I use it now daily to figure out apis, especially when documentation is lacking
20:29:24  <thl0>chrisdickinson: it's a node repl on steroids
20:29:33  <chrisdickinson>thl0: awesome :D
20:30:02  <chrisdickinson>i just recently found out that when you send a USR1 signal to a running node process, you can attach node-inspector to it and live edit the code
20:31:25  <thl0>chrisdickinson: hold on 'reopen the log files' to attach a debugger?
20:31:45  <chrisdickinson>?
20:31:52  <thl0>chrisdickinson: node-inspector needs a lot of love - its almost unusable
20:32:03  <thl0>chrisdickinson: http://stackoverflow.com/questions/5350865/why-do-many-unix-programs-use-signals-like-usr1
20:32:20  <chrisdickinson>yeah, it's a bit out of date, but it seemed to work pretty nicely when i checked last week
20:32:27  <thl0>chrisdickinson: just saw nginx specific
20:32:46  <thl0>chrisdickinson: so USR1 in node means 'attach debugger' ?
20:32:46  <chrisdickinson>thl0: ah, USR1 tells a node process to enable debugging and expose the debug port (5858)
20:32:49  <chrisdickinson>yep
20:32:52  <thl0>cool
20:33:15  <chrisdickinson>the other neat one is that by installing https://npmjs.org/package/heapdump you can make a process generate heapdumps with USR2
20:33:24  <chrisdickinson>which can be loaded in the chrome web dev tools
20:33:27  <thl0>very nice find - this stuff is so powerful especially when combined with replit
20:33:36  * chrisdickinsonnods
20:33:40  <thl0>chrisdickinson: cool
20:33:52  <thl0>chrisdickinson: where is the blog you wrote explaining all this?
20:33:56  <chrisdickinson>haha
20:33:59  <chrisdickinson>in the ether
20:34:03  <chrisdickinson>i need to write up a post
20:34:15  <thl0>:)
20:34:16  <chrisdickinson>i've been just blogging on gists lately
20:34:20  <pkrumins>has anyone used v8-profiler with node-inspector?
20:34:55  <thl0>btw node-inspector hasn't changed in almost a year
20:34:58  <pkrumins>we tried using it yesterday to find the memory leak but it wasn't obvious what to do in the interface in the web node-inspector
20:35:14  <chrisdickinson>ah
20:35:19  <thl0>it works decent, but sometimes break points don't hit which is annoying
20:35:23  <chrisdickinson>pkrumins: were you looking at the heap dumps?
20:35:27  <pkrumins>like there was no profile tab, and sending usr1 to the node processed didnt create one either
20:35:34  <chrisdickinson>right
20:35:42  <pkrumins>are you familiar with that?
20:35:44  <chrisdickinson>you can just open a new tab in chrome and go to the profile tab
20:35:46  <pkrumins>chrisdickinson: i was just running `node --debug server.js`
20:35:49  <chrisdickinson>ahhh
20:36:12  <substack>pending badge statuses work on my local testling build now
20:36:14  <chrisdickinson>pkrumins: USR1 starts debugging a node process
20:36:19  <pkrumins>and then i ran node-inspector on port 9999, and navigated to x.x.x.x:9999/debug?port=5858
20:36:21  <chrisdickinson>so you can point node-inspector at it
20:36:36  <chrisdickinson>if you use heapdump USR2 will create a local heapdump file
20:37:07  <chrisdickinson>download that file, open a tab in chrome, open the profile tab of web tools, right click on the left pane (the list of profiles) and select load
20:37:24  <chrisdickinson>it's a little involved
20:37:39  * thl0is amazed on what's possible if you know your stuff
20:38:22  * mikolalysenkoquit (Ping timeout: 252 seconds)
20:40:39  <pkrumins>chrisdickinson: gotcha
20:41:04  <pkrumins>chrisdickinson: just wondering why the live debugging didnt start via x.x.x.x:9999/debug?port=5858
20:41:16  <pkrumins>that's a bit easier than doing all those tasks
20:41:25  <pkrumins>especially if it's on a remote computer
20:41:40  * douglaslassancejoined
20:42:19  * chrisdickinsonnods
20:43:26  <chrisdickinson>pkrumins: quick screen recording of how to use the heap dump stuff (it's m4v, sorry ): ) http://f.cl.ly/items/2A3T2j411o1m1U1U2t0G/usr2-heap%20-%20Broadband.m4v
20:43:29  <chrisdickinson>thl0: ^^
20:44:05  <pkrumins>looking
20:44:32  <pkrumins>you can just kill -USR2 :)
20:44:43  <pkrumins>chrome plays this nicely
20:45:29  <pkrumins>this is awesome
20:45:40  <pkrumins>chrisdickinson: can you also use it to debug it live?
20:45:43  <pkrumins>without the dump?
20:45:57  <chrisdickinson>pkrumins: you still need to use node-inspector to debug live
20:46:21  <pkrumins>so i'd send USR1 to the node proc, right?
20:46:24  * chrisdickinsonnods
20:46:30  <chrisdickinson>you don't need to start the proc with "--debug" either
20:46:34  <pkrumins>and then what happens, should the profile tab appear suddently?
20:46:44  <chrisdickinson>1 sec, will show :)
20:46:48  <pkrumins>because no matter what i did, i was unable to find the profile tab
20:46:51  <pkrumins>much appreciated
20:52:57  <chrisdickinson>pkrumins: http://f.cl.ly/items/0231212v0J1j0X2g2E2v/usr1%20-%20Broadband.m4v
20:53:05  <pkrumins>watching
20:53:08  <chrisdickinson>so the profile tab isn't in node-inspector, unfortunately
20:53:33  <pkrumins>somehow in felige's tutorial it appears it is: https://github.com/felixge/node-memory-leak-tutorial
20:53:46  <chrisdickinson>huh weird.
20:53:59  <pkrumins>look at that screenshot
20:54:01  <chrisdickinson>yeah
20:54:03  <chrisdickinson>i'll try that out
20:54:12  <pkrumins>yeah i couldn't get it to appear
20:54:14  <chrisdickinson>still, the tooling around v8 is awesome
20:55:22  <thl0>I know no one here cares about CoffeeScript, still I think this is kinda cool: https://github.com/thlorenz/WebGLCraft#coffeescript-support-in-chrome-devtools
20:55:51  * thl0goes into hiding before the haters start throwing stones
21:00:12  <substack>pending statuses work: https://ci.testling.com/substack/node-falafel.png
21:00:34  <substack>click it fast before all the tests finish ^^^
21:02:04  * defunctzombie_zzchanged nick to defunctzombie
21:03:25  <thl0>substack: are you ok with my PR on coffeeify - attaching a root to it?
21:03:48  <substack>I just merged it
21:03:54  <substack>not really sure what it does
21:04:11  <thl0>substack: cool, thanks :)
21:04:12  * substack<-- not a coffee-script user
21:04:26  <substack>thl0: nor are you!
21:04:27  <defunctzombie>substack: nice
21:04:37  <substack>2 non-coffee-script users writing coffee-script tooling
21:05:18  <thl0>substack: no I'm not
21:05:25  * mikolalysenkojoined
21:05:38  * thl0is on a mission to bring even the black sheep into the browserify fold
21:05:56  <substack>it'd be nice if an actual coffee-script user could take over that module
21:06:48  <thl0>substack: agreed - at least now there is a start and almost no friction for the CSers to start using browserify
21:07:32  <thl0>substack: first thing that person would probably do is run the entire project thru jsToCoffee ;)
21:07:42  * thl0cringes
21:08:33  <thl0>substack: time to write a transform that shows browserify bundling progress
21:08:54  <substack>thl0: one thing people are complaining about is how there's no extension api to register .coffee
21:09:23  <substack>https://github.com/substack/node-browserify/issues/326
21:10:44  <thl0>substack: out of scope I think, cause otherwise the next guys are the Roy people who want .roy files to be registered
21:11:01  <thl0>substack: unless there is a generic way to solve this
21:13:03  <substack>transforms can just rewrite require() expressions I figure
21:13:07  <substack>if they really want
21:13:23  <thl0>substack: funny enough that these people are the same that want to edit the bundle ;)
21:14:15  <thl0>substack: agreed
21:15:09  <chrisdickinson>substack: any idea what might be going on here? https://ci.testling.com/chrisdickinson/git-packfile runs locally and passes tests in chrome and firefox using browserify
21:16:05  <chrisdickinson>ah, i bet i know what's going wrong
21:16:11  * mikolalysenkoquit (Ping timeout: 260 seconds)
21:17:45  * ITprojoined
21:17:51  <substack>chrisdickinson: we're still running browserify 1.7 on testling
21:17:58  <chrisdickinson>aaah
21:18:00  <chrisdickinson>that would explain it
21:18:07  <chrisdickinson>D:
21:18:10  <substack>I'm going to overhaul the harnessing soonish
21:18:24  <substack>so that users have control over how bundling happens
21:18:52  <thl0>substack: should we start getting this advanced API for sourcemaps fleshed out and implemented? https://github.com/substack/node-browserify/issues/322#issuecomment-14784651
21:21:51  <substack>I don't really see myself using anything except for inline maps
21:22:07  <substack>it's just too much hassle to put the external map files in the right places
21:23:44  <substack>anyhow tonight I am going to get a system status / logging dashboard running with straggler
21:26:36  <substack>but right now, time for mexican donuts
21:28:24  <thl0>substack: feel the same way - I'm almost tending to tell these people with edgecases to build their own browserify, after all that is what makes modularity shine
21:29:47  <thl0>substack: only thing that would be quite simple and very useful is allowing to override sourceFileRoot to make things appear nicer in devtools
21:31:56  * farnsworthjoined
21:31:57  * cubertjoined
21:35:00  <substack>or a tool that takes browserify bundles on stdin and spits out map files on stdout
21:35:12  <substack>or spits out 2 files: the map and the bundle with the inline maps removed
21:35:15  * marcello3djoined
21:37:07  <thl0>substack: hm, kinda like an after browser-pack transform
21:37:42  <thl0>substack: or something that has nothing to do with browserify at all - got it
21:38:35  <thl0>substack: we should call it deoptimize-source-map ala @domenic's deoptimizer
21:39:55  <thl0>substack: I think that is a great solution, - you can close all these issues and tell them to mod the bundle to their liking
21:40:24  <thl0>substack: it's trivial if they use my module: https://github.com/thlorenz/convert-source-map
21:41:17  <thl0>substack: https://github.com/thlorenz/convert-source-map#fromsource will pull it out of the bundle
21:42:06  * mikolalysenkojoined
21:44:50  * mmckeggquit (Quit: mmckegg)
21:54:42  * mikolalysenkoquit (Ping timeout: 264 seconds)
21:59:46  <thl0>substack: defunctzombie: do you guys have any ideay why my transform only gets called with: bundle-with-progress.js and progressify/index.js ? https://github.com/thlorenz/progressify/blob/master/example/bundle-with-progress.js
22:00:11  <thl0>shouldn't it be called with all the files that get included in the bundle?
22:11:41  * dominictarrjoined
22:14:26  * Domenic_joined
22:20:25  <defunctzombie>pkrumins: for the gif stuff, do you want async api? I meant, it doesn't do any file io
22:20:39  <rvagg>dominictarr, dude, we had to do nodeup without you!
22:20:47  * mikolalysenkojoined
22:21:05  <dominictarr>rvagg: Ah, I figured that out just now
22:21:13  * mmckeggjoined
22:21:24  <rvagg>we waited
22:22:00  <dominictarr>I had expected an email beforehand … I should have waited longer
22:22:02  <dominictarr>sorry
22:23:44  <dominictarr>when is it up?
22:26:10  <rvagg>craig said next couple of days
22:31:32  <dominictarr>does anyone know ho to make a command that auto completes commands in bash?
22:35:27  * mikolalysenkoquit (Ping timeout: 252 seconds)
22:39:20  * st_lukejoined
22:39:22  <pkrumins>defunctzombie: sync is good too
22:39:35  <defunctzombie>k
22:39:55  <pkrumins>node-png is only sycn right now
22:40:16  <pkrumins>we couldn't figure out how to display async updates on the screen in browserling as they come out of order
22:40:22  <dominictarr>isaacs: did you remove the autocomplete feature from npm?
22:40:40  <pkrumins>also it was depressing enough to write the sync part for the new libuv
22:42:32  <rvagg>dominictarr: not sure on mac but on linux is /etc/bash_completion and /etc/bash_completion.d
22:42:54  <rvagg>depending on the dist
22:43:57  <rvagg>pkrumins: perhap the whole uv_poll stuff would help with that
22:44:56  <dominictarr>rvagg: thanks
22:50:48  <pkrumins>rvagg: thanks for the idea
22:50:56  * mmckeggquit (Quit: mmckegg)
22:57:29  <rvagg>pkrumins: these guys do a bit of uv_poll, might be interesting to look at what they're doing: https://github.com/JustinTulloss/zeromq.node/blob/master/binding.cc
22:59:17  * mikolalysenkojoined
23:01:03  * mmckeggjoined
23:06:24  <hij1nx>dominictarr: autocomplete you need to install
23:06:40  <rvagg>yoda talk..
23:06:51  <hij1nx>it simply spits out a string of bash that you can append to your bash profile (or whatever shell you use)
23:06:53  <dominictarr>autocomplete install you must
23:06:53  <rvagg>autocomplete you will have young jedi
23:07:42  <hij1nx>dominictarr: more or less, it works like this https://github.com/hij1nx/complete
23:08:23  <rvagg>is the autocomplete data that works for bash compatible with other shells?
23:10:41  <defunctzombie>isaacs: are native modules built with exceptions disabled now?
23:14:56  * mikolalysenkoquit (Ping timeout: 252 seconds)
23:17:16  <dominictarr>hij1nx: how do I make the example run?
23:17:26  <rvagg>hey dominictarr, hij1nx, or someone, we need a wiki page dedicated to delimeters and characters and whatnot for levelup, that discussion this morning was good but it'd be good if it was written down somewhere
23:17:57  <dominictarr>rvagg: I think I covered that a bit in
23:18:30  <dominictarr>https://github.com/rvagg/node-levelup/wiki/Plugin-Pattern-Discussion#range-bucket
23:18:46  <rvagg>ahhh, excellent
23:18:54  <dominictarr>I should refactor that out, though, since I have changed my approach to level-sublevel
23:22:35  * yorick_quit (Remote host closed the connection)
23:23:45  <hij1nx>dominictarr: what example?
23:23:55  <hij1nx>dominictarr: oh, complete?
23:24:11  <dominictarr>yeah
23:24:21  <dominictarr>https://github.com/hij1nx/complete/issues/13
23:25:06  <hij1nx>dominictarr: just get rid of that from the package
23:25:30  <dominictarr>remove the complete dep?
23:25:32  <hij1nx>dominictarr: is that 0.3.2?
23:25:48  <dominictarr>I cloned from git
23:27:19  <hij1nx>dominictarr: let me take a look
23:27:41  <hij1nx>dominictarr: i just merged a pull request and i thought it all passed, i'll take a look and try it out
23:30:00  <hij1nx>dominictarr: npm install again, i just published
23:30:52  <dominictarr>hij1nx: It should add the exampleapp command right?
23:31:55  <dominictarr>hij1nx: that installed correctly
23:32:56  <hij1nx>dominictarr: what you need to do is make sure you put exampleapp into your path
23:33:10  <hij1nx>in order for it to complete it has to be installed
23:33:24  <hij1nx>give the readme.md a good read, its kind of tricky to do autocomplete
23:33:31  <dominictarr>I npm install -g'd it
23:33:47  <hij1nx>globally installing complete doesnt mater
23:34:23  <hij1nx>oh but i see what you mean, so now it should be in your path because its in the node_modules folder that is already in your path
23:35:16  * mmckeggquit (Read error: Connection reset by peer)
23:36:13  * mmckeggjoined
23:37:06  <hij1nx>dominictarr: you are on osx or linux right?
23:38:47  <hij1nx>heh, actually, i think installing globally will screw some things up
23:39:33  <dominictarr>osx
23:40:17  <substack>I saw some wild chickens in a park wandering around earlier
23:40:20  <hij1nx>ok, so you want to `npm uninstall complete -g`
23:40:40  <hij1nx>dominictarr: installing it globally will override the unix builtin command
23:40:47  <hij1nx>dominictarr: (im guessing)
23:40:56  <hij1nx>dominictarr: you should just install it as a module
23:41:19  * mikolalysenkojoined
23:41:37  <hij1nx>dominictarr: after you install it to your package, put your package in your path
23:42:09  <dominictarr>If I npm install -g it is in my path
23:42:13  <hij1nx>dominictarr: then run your binary that requires `complete`, it will be self explanatory from there on out
23:42:44  <hij1nx>dominictarr: yeah, but `npm install -g complete` will override the built-in unix command that happens to be called `complete`
23:42:55  <hij1nx>dominictarr: there is no reason to install this module globally
23:44:00  <hij1nx>dominictarr: it used to do some fancy global setup stuff where it would figure out your OS and try to append the script to your bash/zsh/fish profile or whatever
23:44:16  <hij1nx>dominictarr: but no self-respecting unix program will write to your profile.
23:44:41  <hij1nx>dominictarr: which is why npm still doesnt do it automatically for you... i guess.
23:45:00  * thl0quit (Remote host closed the connection)
23:45:51  <dominictarr>hij1nx: i didn't npm intstall -g complete in the end,
23:45:58  <dominictarr>just complete/example
23:48:16  <hij1nx>dominictarr: i see, ok so i think i may have recently introduced a bug!
23:48:16  * thl0joined
23:48:31  <hij1nx>dominictarr: or i did not read the last pull request clearly enough
23:49:01  <hij1nx>dominictarr: i can fix this pretty quickly though
23:50:17  * thl0quit (Remote host closed the connection)
23:57:12  * mikolalysenkoquit (Ping timeout: 264 seconds)
23:59:15  * marcello3dquit (Remote host closed the connection)