00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:07  * ircretaryjoined
00:01:19  <Raynos>isaacs, substack: What features does tt lack that tape has?
00:01:31  <isaacs>Raynos: most of them :)
00:01:39  <isaacs>Raynos: tt just wraps require('assert')
00:01:39  <Raynos>:D
00:01:44  <Raynos>I see
00:02:10  <isaacs>Raynos: whereas tape is "node-tap for the browser", tt is "node's builtin assert module, but outputting tap"
00:05:14  * dominictarrquit (Ping timeout: 252 seconds)
00:11:27  * ITproquit (Ping timeout: 260 seconds)
00:16:36  * yorickquit (Remote host closed the connection)
00:21:47  <substack>defunctzombie_zz: "harness": "mocha" is working on my local testling-ci build
00:22:49  <substack>with this all you need to do is add a devDependency for mocha and then set testling.harness to "mocha"
00:23:08  <Raynos>substack: tape closes tape.stream multiple times
00:23:24  <Raynos>because every call to require("tape")(...) will close when that test ends
00:23:52  * ITprojoined
00:31:27  <Raynos>substack: https://github.com/substack/tape/pull/9 fixed
00:31:35  * fotoveritequit (Quit: fotoverite)
00:39:48  <mbalho>isaacs: im guessing this didnt send you emailz https://gist.github.com/4263245
00:39:54  <mbalho>isaacs: (the comments)
00:41:48  <Raynos>mbalho: did you get my emailz
00:42:14  <mbalho>Raynos: today?
00:42:25  <Raynos>from the comment.
00:42:26  <Raynos>on that gist
00:42:32  <mbalho>nope
00:42:34  <mbalho>still broken
00:42:36  <mbalho>damn github
00:47:16  <isaacs>mbalho: nope
00:47:20  <isaacs>mbalho: no emails
00:48:42  <mbalho>GITHUB: Y U NO TELL ME THINGS
00:52:41  <isaacs>i actually love not getting emails on gists.
00:52:46  <isaacs>that's kind of brilliant
00:54:28  <mbalho>wat
00:54:33  <mbalho>i want to know when people ask me things
00:54:36  <mbalho>or at least have a option
00:55:41  * saijanai_joined
00:57:54  * _sorensenquit (Ping timeout: 240 seconds)
00:59:02  <Raynos>substack: sync race condition https://github.com/substack/tape/pull/10
00:59:49  <substack>testling-ci mailing list announcement post https://gist.github.com/raw/c1ea94505e32fcec689e/58f5b506d0606364f3261f2320045649af8db904/gistfile1.txt
00:59:56  <substack>feedback welcome before I post it for realz
01:05:01  <substack>posting in 3 2
01:05:13  <substack>posted
01:16:10  <Raynos>substack: I like how your actually branding it as the PROFESH edition
01:16:13  <Raynos>and it's not an internal joke.
01:18:22  <jjjjohnnny>looks good
01:23:00  * fotoveritejoined
01:23:26  <fotoverite>yo anyone know tarr's number?
01:24:53  * fotoveritequit (Client Quit)
01:28:13  <Raynos>fotoverite: dont think he has a numer
01:32:44  <substack>Raynos: I dislike corporate speak and suspect that most developers feel likewise
01:32:49  <substack>or most humans for that matter
01:33:15  <Raynos>Oh shit I forgot to add tests to my pull requests >_<
01:38:07  * thatguydanjoined
02:13:53  <Raynos>substack: trust me on these: https://github.com/substack/tape/pulls
02:14:28  <substack>on it
02:15:33  <substack>blarg your editor "fixes" leading whitespace justified to the current indentation block
02:15:36  <substack>noisy
02:16:10  <substack>I really want a tool for managing github pull requests without having to open up the github ui hmmm
02:16:44  <substack>Raynos: your counter thing seems to break the test/harness.js test
02:17:39  <substack>patch 3 conflicts with patch 4
02:19:54  <Raynos>substack: >_<
02:20:06  <Raynos>Ok
02:20:10  <Raynos>I will go and fix the tests properly
02:21:21  <Raynos>substack: run tape tests on ci.testling!
02:25:59  <substack>yes but
02:26:10  <substack>I need a thing to let me output tap for the tape tests!
02:31:28  <Raynos>xd
02:31:33  <Raynos>use tape to test tape
02:31:38  <Raynos>Oh no wait
02:31:42  <Raynos>use isaacs/tt!
02:33:17  <Raynos>ok I found the bug
02:33:54  <Raynos>https://github.com/substack/tape/pull/9 bug free
02:34:42  * jiang-plusjoined
02:38:22  <dools>substack: just use tape to output tap for the tape tests
02:38:57  <dools>the likelihood that you'd have a bug in tape that meant it was capable of running, and producing incorrect output that appeared to be correct, is pretty slim
02:40:23  <Raynos>substack: https://github.com/substack/tape/pull/10
02:40:31  <Raynos>I merged #9 into #10
02:52:39  * jiang-plusquit (Quit: Lost terminal)
03:13:24  * ralphtheninjaquit (Ping timeout: 248 seconds)
03:28:58  * defunctzombie_zzchanged nick to defunctzombie
03:30:42  * ITproquit (Ping timeout: 264 seconds)
03:34:08  <defunctzombie>substack: any reason you went for testling info in package.json vs .testling.yml or such?
03:35:02  <defunctzombie>substack: also, setting up int for testling, I think it would be useful to "blacklist" browsers vs whitelist
03:35:09  <defunctzombie>that list would be smaller
03:35:22  <defunctzombie>basically, I want to say test in all but these
03:35:30  <defunctzombie>so that I automatically get any new browsers you have
03:37:03  <defunctzombie>cause keeping up with wahtever is the "latest" version of these browsers would be annoying
03:39:24  <substack>defunctzombie: true but it would be kind of a waste to run tests in all the browsers since there are so many
03:39:41  <substack>perhaps a "latest"/"latest-N" syntax?
03:39:45  <defunctzombie>yea
03:39:48  <defunctzombie>that would be good too
03:39:51  <defunctzombie>I tried chrome/latest
03:39:54  <defunctzombie>but that didn't work
03:40:04  <defunctzombie>http://ci.testling.com/shtylman/node-int
03:40:19  <defunctzombie>also, the test seemed to fail due to undefined test variable
03:40:26  <defunctzombie>not sure if that is something I need to fix or what
03:40:50  <substack>looking
03:41:12  <defunctzombie>I guess for the browsers the reason I want to blacklist is because I really want to make it works in all the various versions
03:41:19  <defunctzombie>and only ignore the ones I know I don't want to support
03:41:29  <defunctzombie>otherwise I will constantly be adding new versions as they are released
03:41:41  <defunctzombie>and if I use only chrome/latest then I stop testing older versions
03:41:46  <defunctzombie>when you guys bump the chrome version
03:41:53  <defunctzombie>which is not what I want to happen
03:41:56  <substack>chrome/canary is the nightly you can do
03:42:09  <substack>agreed about "latest" though
03:42:10  <defunctzombie>(thinking about it from a standpoint that users will have all sorts of versions)
03:42:52  <substack>defunctzombie: oh you're running mocha in tdd mode
03:42:59  <substack>I have mocha configured to do bdd mode
03:43:02  <defunctzombie>mostly because now I need to think about.. ok.. so chrome/23 came out do I update my package file
03:43:09  <defunctzombie>substack: ah, I hate bdd mode ;)
03:43:12  <substack>same
03:43:19  <substack>blarg looks like I need to read the mocha.opts
03:43:26  <defunctzombie>:D
03:43:37  <substack>https://github.com/shtylman/node-int/blob/master/test/mocha.opts
03:43:40  <substack>what format is this even
03:43:44  <substack>shell opts?
03:44:17  <defunctzombie>I think so
03:44:21  <defunctzombie>ghetto ass format
03:44:23  <defunctzombie>is the format
03:44:34  <substack>how about
03:44:37  <substack>mocha-THING
03:44:42  <substack>in the "harness" field
03:44:46  <substack>mocha-qunit in your case
03:44:56  <substack>since the reporter is always "tap"
03:45:02  <defunctzombie>would be nice if it "just worked"
03:45:09  <defunctzombie>but I can compromise :)
03:45:25  <defunctzombie>duplication means room for mistakes, but they should be obvious quickly enough
03:45:25  <substack>mocha sort of deserves to not work
03:45:29  <substack>so much magic
03:45:39  <defunctzombie>haha, I almost agree
03:45:43  <defunctzombie>I was gonna switch to tape
03:45:51  <substack>well if you already have working tests
03:45:56  <defunctzombie>yep
03:46:12  <defunctzombie>for tape I just needed better exception stacks
03:46:19  <defunctzombie>server side, otherwise debug is a bitch
03:46:27  <substack>yep
03:46:32  <defunctzombie>http://ci.testling.com/shtylman/node-int
03:46:36  <defunctzombie>so yea, I added the other browsers
03:46:50  <defunctzombie>def would be nice to have the blacklist or some way to not have to think about this as much :/
03:47:10  <defunctzombie>cause I have no clue what the latest in each is or latest - 1 or shit like that
03:47:28  <defunctzombie>and yea, I can see how it can be a chore to run in lots of browsers
03:47:36  <substack>http://browserling.com/browsers.json
03:47:38  <defunctzombie>but that does seem like part of the point ;)
03:48:23  <defunctzombie>so what is the word on the tdd stuff? mocha-tdd field value?
03:48:28  <defunctzombie>or detected from mocha opts?
03:48:40  <substack>I'm leaning towards mocha-tdd
03:48:51  <substack>mocha.opts looks like a thing that is subject to change/removal
03:52:34  <defunctzombie>k
03:52:38  <defunctzombie>fair enough
03:53:50  <defunctzombie>let me know whenever you want me to try that out (no rush ;P)
03:54:05  <substack>hacking in mocha-$UI mode right now
03:54:10  <substack>but it won't be up until tomorrow
03:54:39  <defunctzombie>kk
04:04:47  <substack>ok mocha-tdd mode works in my local build
04:04:58  <defunctzombie>coo
04:05:09  <substack>documenting
04:05:17  <substack>this stuff will be live when pkrumins wakes up
04:05:42  <defunctzombie>\o/
04:14:57  * tphummelquit (Quit: tphummel)
04:19:44  <Raynos>substack: http://www.youtube.com/watch?v=CIY6qNk1zy0&feature=plcp
04:19:48  <Raynos>^ that is badass
04:22:35  <substack>paul_irish: thanks for the g+ reshare \o
04:23:04  <substack>once mocha-$UI goes live tomorrow testling-ci will work with qunit tests via "harness": "mocha-qunit"
04:23:48  <Raynos>whats mocha ui?
04:24:06  <Raynos>is that the mootools gui?
04:24:32  <substack>mocha --ui $UI
04:24:39  <substack>mocha --interfaces gives the complete list
04:24:47  <substack>bdd, tdd, qunit, exports
04:24:48  <Raynos>Oh
04:25:06  <Raynos>we should build that on top of tape :P
04:25:19  <Raynos>I can totally hack all of those interfaces on top of tape :D
04:35:01  <substack>does the github download feature removal break componentjs?
04:35:03  <substack>I hope so
04:36:28  <substack>oh it looks different
04:55:52  * dominictarrjoined
05:14:14  <defunctzombie>the problem I have with these non-refresh things at the moment is that they don't handle complex app logic well, when I refresh changes, they are not because I don't know what a variable change will do. It is usually because I need to rerun a lot more code from the start
05:15:45  * antixjoined
05:15:45  * antixquit (Changing host)
05:15:45  * antixjoined
05:18:30  <Raynos>defunctzombie: psh.
05:18:34  <Raynos>build your app
05:18:40  <Raynos>so that it can be restarted arbitarly
05:25:43  <jjjjohnnny>HRM HRM AND A HO HO WHAT WHAT
05:26:24  <jjjjohnnny>GUTEN IDEA / GUTEN-FREE IDEA?
05:27:44  <jjjjohnnny>CITIZEN AS A SOFTWARE U SAY?
05:29:17  <defunctzombie>hahaha
05:29:46  <defunctzombie>o.O
05:37:35  * st_lukejoined
05:48:13  * st_lukequit (Remote host closed the connection)
06:03:27  * fotoveritejoined
06:25:36  <substack>bwahaha I have something much better than bouncy
06:25:39  <substack>approach-wise
06:25:43  <substack>so simple
06:28:45  <substack>and faster than http-proxy
06:31:27  <substack>well it does less so I should hope so
06:32:57  <substack>crap the memory usage keeps going up though
06:33:03  <substack>oh wait there it got GC'd
06:34:17  <Raynos>i got my p2p chat demo for thursday working!
06:34:19  <Raynos>exciting times
06:34:25  <Raynos>now to refactor it to use less abstractions
06:34:29  <Raynos>and inline all the DOM bullshit!
06:35:40  <substack>oh crap this thing is eating memory like crazy
06:53:17  <Raynos>http://raynos.github.com/painter/examples/chat/static/
06:53:19  <Raynos>p2p chat!
06:57:52  * b__joined
07:14:15  * b__quit (Quit: Computer has gone to sleep.)
07:25:19  <jjjjohnnny>Raynos: I cant click on Message to send a message
07:25:27  <Raynos>type a name
07:25:29  <Raynos>hit enter
07:32:55  <jjjjohnnny>Raynos: where is bundle js?
07:32:59  <jjjjohnnny>in the repo
07:33:06  <Raynos>in the gh-pages branch
07:33:12  <Raynos>its auto generated by browserify
07:33:29  <Raynos>https://github.com/Raynos/painter/blob/master/examples/chat/Makefile#L1
07:33:36  <Raynos>you dont want to read the bundle its noise
07:34:00  <jjjjohnnny>right
07:34:16  <jjjjohnnny>but this is psuedo p2p correct?
07:35:25  <Raynos>yes.
07:35:56  <Raynos>until chrome 25
07:36:00  <jjjjohnnny>is this using "fully connected"
07:36:01  <Raynos>which should have working datachannels
07:36:16  <Raynos>the fully connected network topology where everyone connects to everyone
07:36:54  <jjjjohnnny>why not use star?
07:37:08  <jjjjohnnny>would that work just the same in this case?
07:37:20  <Raynos>its more complicated to implement
07:37:28  <Raynos>because you need to handle the case where the center of the star goes down
07:37:43  <Raynos>i.e. when the network partitions you need to open more connections to make sure that no nodes are isolated
07:38:16  <Raynos>but this code does not relay or gossip chat messages
07:38:37  <Raynos>so it each message send to a peer only goes one hop and would never get further so wont work with a star network
07:38:37  <jjjjohnnny>what does it do
07:38:50  <Raynos>because it would have to go two hops. client -> center of start -> other client
07:39:09  <Raynos>you should come to the meetup on thursday
07:39:11  <Raynos>and ask this question :P
07:39:22  <jjjjohnnny>yes.. but... that is what it's doing o.O
07:39:30  <jjjjohnnny>:p
07:40:03  <jjjjohnnny>i'll show up but I aint on the list
07:40:38  <Raynos>its cool :D
07:40:46  <Raynos>probably >_>
07:40:58  <Raynos>you should be able to hustle your way in
07:41:45  <jjjjohnnny>"im not on the list, but I am a developer looking for a nice place to work"
07:42:30  <Raynos>:D
07:42:39  <Raynos>and your one of my guests
07:52:04  * tphummeljoined
07:56:12  * tphummelquit (Client Quit)
07:58:36  * defunctzombiechanged nick to defunctzombie_zz
07:58:57  * slaskisjoined
08:05:27  * wiwilliajoined
08:11:23  * thatguydanquit (Quit: thatguydan)
08:13:11  * b__joined
08:26:34  * thatguydanjoined
08:48:31  * wiwilliaquit (Ping timeout: 260 seconds)
09:25:41  * thatguydanpart
09:26:57  * ITprojoined
09:31:24  * ITproquit (Ping timeout: 264 seconds)
09:32:14  * jez0990_joined
10:13:09  * b__quit (Quit: Textual IRC Client: www.textualapp.com)
10:26:00  * jez0990_quit (Remote host closed the connection)
10:35:10  * thatguydanjoined
10:35:12  * thatguydanquit (Client Quit)
10:35:42  * b__joined
10:36:39  * slaskisquit (Quit: slaskis)
10:48:21  * farnsworthquit (Ping timeout: 245 seconds)
10:48:21  * cubertquit (Ping timeout: 245 seconds)
10:52:51  * ITprojoined
10:52:58  * farnsworthjoined
10:53:08  * ITproquit (Client Quit)
10:53:59  * cubertjoined
11:16:58  * slaskisjoined
11:20:17  * jez0990joined
11:24:13  <slaskis>substack: so, should this work now? https://github.com/publicclass/geom-vec/blob/master/package.json
11:28:12  <substack>slaskis: yep, try it!
11:28:39  <slaskis>substack: there's nothing here now though: http://ci.testling.com/publicclass/geom-vec is there anything else i have to do?
11:28:55  <substack>slaskis: did you add the admin hook in the github ui?
11:29:27  <slaskis>i did (but double checking now)
11:29:43  <substack>and then you need to make a new commit to the repo or send a test hook if you haven't made any commits
11:29:49  <substack>since adding the webhook
11:30:06  <slaskis>i sent a test hook just now
11:30:46  <slaskis>it was there since i started bugging you with mocha support ;)
11:31:18  <substack>you might need to make a new commit
11:32:01  <slaskis>ok, lets see..
11:33:47  <slaskis>unrelated to my repo it seems like your static file module confuses project.js with project.css some times, and i get 404s for them too some times
11:34:07  <substack>that isn't the issue
11:34:14  <substack>the issue is that the http router is buggy as fuck
11:34:19  <slaskis>oh ok
11:34:28  <substack>crashes all the time
11:34:32  <substack>working on the replacement
11:34:45  <slaskis>ah. what are you using now? (so i can avoid it)
11:35:03  <substack>bouncy
11:35:07  <substack>which I wrote
11:35:11  <slaskis>oh i use that one haha
11:35:13  <substack>the next version will be much better
11:35:24  <slaskis>still nothing on http://ci.testling.com/publicclass/geom-vec
11:35:41  <slaskis>is there a log ?
11:35:54  <substack>not anything I can easily get at
11:35:58  <substack>sleeping here
11:36:00  <substack>super late
11:36:08  <slaskis>oh right
11:36:14  <slaskis>i won't keep you up then
12:29:33  * b__quit (Quit: Textual IRC Client: www.textualapp.com)
12:44:27  * ralphtheninjajoined
13:38:14  * farnsworthquit (Ping timeout: 240 seconds)
13:39:30  * farnsworthjoined
14:09:02  * dominictarrquit (Read error: Operation timed out)
14:11:54  * juliangruber_joined
14:12:07  * jez0990quit (Ping timeout: 252 seconds)
14:19:38  * iancrowtherjoined
14:20:30  <iancrowther>Hey, so I got here from http://ci.testling.com/ and am wondering what is involved in adding mobile browsers to this stack
14:24:42  * jdenquit (Ping timeout: 259 seconds)
14:24:51  * jdenjoined
14:25:15  * jdenchanged nick to Guest96841
14:26:25  * duncanbeeverschanged nick to birdbones
14:27:13  * b__joined
14:30:47  * iancrowtherquit (Quit: Page closed)
15:00:53  * ralphtheninjaquit (Ping timeout: 252 seconds)
15:25:14  * juliangruber_quit (Quit: juliangruber_)
15:31:01  * juliangruber_joined
15:40:06  * defunctzombie_zzchanged nick to defunctzombie
15:42:25  <juliangruber_>substack pkrumins: mobile browsers would rock!
15:53:11  * dominictarrjoined
16:01:07  * juliangruber_quit (Ping timeout: 260 seconds)
16:02:22  * b__quit (Quit: Computer has gone to sleep.)
16:17:26  * b__joined
16:17:58  * wiwilliajoined
16:21:23  * AvianFluquit (Remote host closed the connection)
16:50:58  * juliangruber_joined
16:52:28  * juliangruber_quit (Client Quit)
16:53:42  * st_lukejoined
17:04:16  * _ritchjoined
17:05:40  <rowbit>/!\ ATTENTION: (default-local) denni@... successfully signed up for developer browserling plan ($20). Cash money! /!\
17:05:40  <rowbit>/!\ ATTENTION: (default-local) paid account successfully upgraded /!\
17:10:40  * jez0990_joined
17:13:55  * AvianFlujoined
17:22:05  <defunctzombie>what does it mean when there is no entry under a browser? is the test stalled? http://ci.testling.com/shtylman/node-int
17:29:35  <st_luke>you broke it
17:30:00  <defunctzombie>fuck
17:31:18  <pkrumins>looking
17:31:31  * defunctzombiechanged nick to defunctzombie_zz
17:31:40  * b__quit (Quit: Computer has gone to sleep.)
17:33:35  <pkrumins>we'll try to reproduce this, i'll just clone your repo and push some changes
17:40:45  <slaskis>pkrumins: can you see if testling ci gets anything from publicclass/geom-vec ?
17:41:42  <pkrumins>looking
17:41:56  <pkrumins>yes i'm seeing it
17:42:00  <pkrumins>npm failed
17:42:13  <pkrumins>here is the reason: npm ERR! 404 'geom-mat' is not in the npm registry.
17:42:29  <pkrumins>so npm failed and didnt proceed to run any tests as installing geom-mat failed
17:43:25  <slaskis>ooh
17:43:27  <slaskis>right
17:43:29  <slaskis>thanks
17:45:33  * tphummeljoined
17:56:48  <dominictarr>substack, I'm trying to figure out how to change map-reduce so it has a combiner
18:00:57  <substack>I was looking at that
18:01:34  <dominictarr>the problem is that the user knows better what the top level reduce is than the module
18:02:16  <dominictarr>if they always used a strict number of groups, that would work.
18:02:49  <dominictarr>because I'd just check whether the key array was a given length
18:03:47  * b__joined
18:04:30  <dominictarr>I'm not sure that there are any use cases where you might want different groups
18:05:00  <substack>different groups?
18:05:24  <dominictarr>emit([this, that], value); emit(['anotherThingEntirely'], value2)
18:05:56  <dominictarr>but that is weird.
18:06:00  <dominictarr>hey
18:06:09  <dominictarr>relocating
18:06:10  <dominictarr>brb
18:06:54  <substack>if there is no combine field it can work the way it currently does
18:07:08  <substack>but if there is a combine field, just put the rereduces through that instead of reduce
18:10:43  * dominictarrquit (Ping timeout: 246 seconds)
18:14:27  <slaskis>substack: thanks to pkrumins i got it up and running! http://ci.testling.com/publicclass/geom-vec
18:14:33  <slaskis>although everything fails
18:17:07  <pkrumins>nice!
18:19:07  <slaskis>does testling add a NODE_ENV=test env var? so i can remove some noise in the tests...
18:19:39  <pkrumins>it doesn't
18:22:26  <slaskis>i guess i could just set it in my test.js too...
18:39:49  <chrisdickinson>so, dumb browserify use case -- not for me, but for someone in #webgl. they're claiming that they can't use it because they use their own handrolled system to pull down a tarfile full of game assets and then mount that as a "virtual filesystem". i think browserify can handle this use case (at least through linking in those assets at compile time), but is there any machinery in browserify to allow for runtime definition of modules?
18:40:20  <chrisdickinson>"virtual filesystem" are his words, not mine.
18:42:06  * dominictarrjoined
18:49:37  <jjjjohnnny>either substack's alarm is going off in his room or he is practicing his frequency ululations
18:50:57  <jjjjohnnny>humna humna humna
18:51:12  <substack>chrisdickinson: tell him his release management is really poor
18:52:29  <jjjjohnnny>stackVM ringtones
18:55:12  <jjjjohnnny>the king kong who sing songs and made a million dollars on ring tones
18:55:33  * ralphtheninjajoined
18:55:42  <chrisdickinson>substack: yeahhh. trying to nudge him onto a nicer course, since he's got really nice webgl demos and it'd be cool to have them available via npm.
18:56:00  <chrisdickinson>so there's no way to define modules at runtime?
18:56:57  <substack>chrisdickinson: there's a "scripts" field
18:57:03  <substack>for people who don't use commonjs
18:58:23  <chrisdickinson>hmm. so it's likely that some of this tarball contains packaged assets -- images, shaders, etc.
18:58:37  <chrisdickinson>sorry if this is a weird use case.
19:03:49  <ralphtheninja>7
19:03:52  <ralphtheninja>ups
19:04:49  * slaskisquit (Quit: slaskis)
19:05:43  <substack>chrisdickinson: another thing is that testling-ci does `npm install .` so you can put custom asset fetching steps in the scripts fields
19:06:33  <chrisdickinson>substack: ooc, is there any support for stubbing out XHRs in testling or would that be the responsibility of the test author?
19:06:55  <chrisdickinson>(i'm assuming the latter.)
19:07:19  <substack>yes that is the responsibility of the test author
19:07:50  <substack>at some point later I want to let people spin up their apps completely to run tests against living servers
19:10:40  <substack>dominictarr: http://www.newscientist.com/blogs/culturelab/2012/12/momath-manhattan-museum-of-mathematics.html
19:10:46  <substack>opens on the 15th!
19:11:02  <Raynos>substack: randpeer video went missing
19:11:11  <Raynos>:(
19:11:20  <dominictarr>oh, sweet!
19:11:21  <substack>hij1nx, fotoverite, defunctzombie_zz too: ^^^ math museum woo
19:11:51  <fotoverite>yup
19:11:58  <fotoverite>I think I blog about that.
19:12:09  <fotoverite>Def up for it.
19:12:09  <substack>saw it on twitter
19:12:52  * defunctzombie_zzchanged nick to defunctzombie
19:13:14  <defunctzombie>substack: yea, I can't wait to go :)
19:14:02  <substack>brianloveswords too!
19:14:11  <substack>and any other NYers I may have missed
19:14:21  <defunctzombie>haha
19:15:18  <chrisdickinson>it might be cool to have a series of modules that stub out things like XHR, websockets, eventsource, images and friends and make them point at a local, in-browser fake "server".
19:15:20  <st_luke>I'm gonna try something new with recruiters that are spamming me lately
19:16:25  <st_luke>going to try going resume-less
19:17:04  <defunctzombie>substack: I think I made testling stall
19:17:16  <defunctzombie>http://ci.testling.com/shtylman/node-int
19:17:22  <defunctzombie>there is nothing underfirefox
19:17:30  <defunctzombie>and ie8 seems to have... hm.. not finished
19:18:06  <substack>st_luke: oh whoops forgot you are in NY!
19:18:11  <pkrumins>defunctzombie: http://ci.testling.com/pkrumins/node-int
19:18:26  <pkrumins>defunctzombie: it was a browser glitch of some kind
19:18:33  <pkrumins>wanted to reproduce that bug but it worked fine
19:19:01  <pkrumins>i'll try to catch it when it happens again
19:19:07  <substack>chrisdickinson: I think it's better to test against the real thing
19:20:03  * b__quit (Quit: Computer has gone to sleep.)
19:20:50  <defunctzombie>pkrumins: huh, interesting. and the ie8 failure? it seems to not finish
19:21:00  <substack>well that's IE for you
19:21:05  <defunctzombie>:D
19:21:15  <substack>possibly crashed where it left off
19:21:32  <defunctzombie>hahaha
19:21:33  <defunctzombie>fucking ie
19:21:50  <pkrumins>oh cool discovery
19:21:58  <pkrumins>looking at what's happening to ie when it runs the test
19:22:00  <defunctzombie>if I hit the webhook url with the right stuff, can I make it re-run the test?
19:23:40  <pkrumins>oh i see
19:23:44  <pkrumins>IE didnt crash
19:24:13  <pkrumins>it just said - this script is causing the page to load slowly, do you wish to continue running it?
19:24:17  <pkrumins>with a yes and no buttons
19:25:24  <pkrumins>defunctzombie: not right now
19:27:43  <pkrumins>defunctzombie, substack: http://i.imgur.com/d1bJx.png
19:28:16  <defunctzombie>hahaha
19:28:35  <pkrumins>deploying a fix http://support.microsoft.com/default.aspx?scid=kb;en-us;175500
19:29:03  <defunctzombie>wow
19:29:42  <defunctzombie>IE... what a shitshow
19:30:15  <pkrumins>why? all browsers do that
19:30:23  <pkrumins>it just had a lower limit
19:30:24  <defunctzombie>pkrumins: thanks for looking into this :)
19:30:40  <pkrumins>ok fix deployed. testing it out.
19:30:49  <pkrumins>sure
19:32:34  <pkrumins>defunctzombie, substack, http://ci.testling.com/pkrumins/node-int <- IE8 fixed
19:33:01  <substack>\o/
19:33:20  <pkrumins>now chrome has the same issue
19:33:24  * shuaibjoined
19:33:54  <pkrumins>safari too
19:36:51  <pkrumins>i see the VM was slow
19:36:56  <pkrumins>so chrome and safari
19:36:58  <pkrumins>boring
19:37:49  <pkrumins>so chrome and safari didnt finish running tests before timeout kicked in and they got killed
19:38:31  <defunctzombie>ah
19:42:24  * slaskisjoined
19:42:53  <pkrumins>oh actually it might have been a bug
19:43:13  <defunctzombie>bug in the module or test env?
19:44:31  <pkrumins>bug in testling
19:44:33  <hij1nx>dominictarr: http://www.yelp.com/biz/milk-and-roses-brooklyn
19:44:41  <dominictarr>fotoverite, ^
19:45:02  <hij1nx>fotoverite: ^^ :)
19:45:11  <fotoverite>I see what time?
19:45:20  <hij1nx>what sounds good to you?
19:45:30  <hij1nx>were coding at some cafe in manhattan atm
19:45:36  <fotoverite>I can do in probably an hourish
19:46:41  <fotoverite>What the hell did I just watch
19:46:42  <fotoverite>http://www.giantbomb.com/metal-gear-solid-v-or-the-phantom-pain-or-whatever-theyll-eventually-call-it-debut-trailer/17-6879/
19:55:50  * _sorensenjoined
20:00:35  <CoverSlide>a phantom game with the phantom name
20:05:13  <Raynos>substack: https://github.com/Colingo/valid-schema
20:05:17  <Raynos>testling CI is pretty badass
20:06:40  <pkrumins>awesome
20:07:26  <dominictarr>mbalho, oh I was trying to use gather yesterday
20:07:49  <dominictarr>had a problem because I had not set my local time to eastern yet
20:08:31  <dominictarr>you should check whether the time is the right timezone for the location.
20:08:45  <Raynos>substack: http://ci.testling.com/Colingo/valid-schema
20:08:58  <Raynos>that page shows nothing
20:09:27  <pkrumins>shows everything here
20:13:16  <Raynos>I see nothing
20:13:23  <Raynos>oh now I see it
20:15:05  * st_luke_joined
20:15:36  * st_lukequit (Read error: Connection reset by peer)
20:16:00  <defunctzombie>these references to global in engine.io are completely silly
20:16:46  <Raynos>defunctzombie: in engine.io or engine.io-client?
20:16:51  <defunctzombie>client
20:16:56  <defunctzombie>sorry, wasn't clear
20:20:32  <Raynos>defunctzombie: well engine.io-client already works with browserify so i dont care :D
20:20:50  <defunctzombie>at least you consolidated them to one place
20:21:06  * ralphtheninjaquit (Ping timeout: 244 seconds)
20:21:10  <defunctzombie>Raynos: I know you don't care, this isn't for you
20:21:21  <defunctzombie>I care when I see something that I can think I can improve ;)
20:21:55  <hij1nx>defunctzombie: Raynos: I think im going to make a "sparse" fork of engine.io
20:22:18  <defunctzombie>meaning?
20:22:22  * b__joined
20:23:09  <hij1nx>defunctzombie: current/head version is trying to make it `component` enabled. so it conflates build concepts with run-time code. dont like that (hence the globals and other silly things)
20:23:43  <hij1nx>but overall i like it. but i think it could be a lot simpler.
20:23:55  <defunctzombie>hij1nx: I have one that doesn't confuse the two :)
20:24:00  <defunctzombie>that is what I have been complaining about
20:24:10  <hij1nx>defunctzombie: link?
20:26:27  <defunctzombie>https://github.com/shtylman/engine.io-client/commits/browser
20:26:27  <defunctzombie>working on rebasing on master right now
20:26:27  <defunctzombie>almost done
20:26:27  <defunctzombie>basically, what this does it use all the modules as you would expect
20:26:27  <defunctzombie>and allows the bundler/tool to pickup on a "browser" field in package.json
20:26:27  <defunctzombie>(which I have made teh xmlhttprequest and ws modules provide)
20:26:27  <defunctzombie>so when you create package for the browser, it just does the right thing
20:26:29  <defunctzombie>without having to put build shit and runtime checks in the code
20:28:35  <defunctzombie>ok, rebased on master
20:30:51  * ITprojoined
20:31:06  <defunctzombie>hij1nx: take the branch of engine.io-client I linked and use try this https://github.com/shtylman/node-script
20:31:29  <defunctzombie>install script from github for now, as I haven't published hte latest versions
20:31:56  <defunctzombie>the effect should be that it detects the deps for engine.io-client properly and packages engine.io-client accordingly
20:32:15  <defunctzombie>my end goal is to remove all the references to global from engine.io-client but I haven't gotten there yet
20:43:19  * juliangruber_joined
20:43:23  * juliangruber_quit (Read error: Connection reset by peer)
20:44:10  * juliangruber_joined
20:44:17  * juliangruber_quit (Read error: Connection reset by peer)
20:46:18  * juliangruber_joined
20:46:26  * juliangruber_quit (Read error: Connection reset by peer)
20:48:29  * juliangruber_joined
20:48:37  * juliangruber_quit (Read error: Connection reset by peer)
20:50:09  * elliottcablequit (Quit: Lost terminal)
20:50:09  * purrquit (Read error: Connection reset by peer)
20:50:18  <dominictarr>https://github.com/dominictarr/scuttlebucket/
20:50:40  * juliangruber_joined
20:50:44  <substack>haha yes
20:50:48  * juliangruber_quit (Read error: Connection reset by peer)
20:51:38  <substack>dominictarr: I think your scuttlebutt tests in testling ci are failing because of process.on('exit', fn)
20:52:48  * juliangruber_joined
20:53:12  <defunctzombie>the fact that node made process a global is sad :( would have been better as a module
20:53:12  * juliangruber_quit (Read error: Connection reset by peer)
20:53:45  <pkrumins>dominictarr: s/https/http/
20:53:46  * juliangruber_joined
20:53:51  * juliangruber_quit (Read error: Connection reset by peer)
20:54:37  * juliangruber_joined
20:54:42  * dominictarrquit (Ping timeout: 244 seconds)
20:54:45  * juliangruber_quit (Read error: Connection reset by peer)
20:55:29  * juliangruber_joined
20:55:38  * juliangruber_quit (Read error: Connection reset by peer)
20:56:05  * tilgovijoined
20:56:22  * juliangruber_joined
20:56:28  * juliangruber_quit (Read error: Connection reset by peer)
20:58:29  * juliangruber_joined
20:58:42  * juliangruber_quit (Read error: Connection reset by peer)
21:00:40  * juliangruber_joined
21:00:48  * juliangruber_quit (Read error: Connection reset by peer)
21:02:51  * juliangruber_joined
21:02:56  * juliangruber_quit (Read error: Connection reset by peer)
21:04:36  * juliangruber__joined
21:04:40  * juliangruber__quit (Read error: Connection reset by peer)
21:04:46  * fotoveritequit (Quit: fotoverite)
21:06:46  * juliangruber_joined
21:06:52  * juliangruber_quit (Read error: Connection reset by peer)
21:07:37  * juliangruber_joined
21:07:44  * juliangruber_quit (Read error: Connection reset by peer)
21:09:46  * juliangruber_joined
21:09:54  * juliangruber_quit (Read error: Connection reset by peer)
21:11:00  * ITproquit
21:11:24  * ITprojoined
21:13:41  <substack>exciting stuff happening yay
21:18:09  <defunctzombie>The cafe I am at has flaky internet... this is quite annoying #programmerproblems
21:21:31  * defunctzombiereally wants to find a good use for scuttlebutt stuff
21:23:36  <substack>are you using redis for anything?
21:23:53  <substack>or storing ephemeral state data in a db
21:24:02  <substack>those are the best immediate uses for it for most apps
21:24:43  <st_luke_>defunctzombie: let me know if you come up with anything interesting
21:25:17  <defunctzombie>hm, not really atm
21:25:25  <st_luke_>I want to find a good excuse to spin up a bunch of VMs at the data centers in various continents before my access gets cut off
21:25:57  <slaskis>substack: are the errors reported at http://ci.testling.com/publicclass/geom-vec my js or can it have been an old bug in testling? because the lib passes the tests locally...
21:26:16  <pkrumins>looking
21:26:42  <substack>slaskis: is b9571a3f141fc1fba29c01f9fda9370c84cc5be5 the latest commit locally?
21:26:44  * jez0990_quit (Ping timeout: 248 seconds)
21:26:50  <slaskis>yep
21:26:54  <substack>hmm
21:27:01  * defunctzombiechanged nick to defunctzombie_zz
21:27:12  <pkrumins>slaskis: isnt it related to that module?
21:27:13  <substack>slaskis: here's what your tests compiled down into http://git.testling.com/bundle/1355336053729.1df71f84.js
21:27:27  <pkrumins>that i said was missing?
21:27:31  <st_luke_>defunctzombie_zz: singapore, austin, amsterdam, san jose, nyc. would be perfect for something fun/distributed
21:27:33  * elliottcablejoined
21:27:35  <pkrumins>oh ok if tests compiled then all modules were present
21:27:58  * purrjoined
21:28:11  <pkrumins>slaskis: looking in the logs
21:28:17  <substack>slaskis: seems like there was a syntax error in the index.js file
21:28:20  <pkrumins>to see if i can find the log of your test
21:28:21  <slaskis>pkrumins: thanks :)
21:28:49  * st_luke_changed nick to st_luke
21:28:56  <pkrumins>slaskis: no log of that
21:29:01  <pkrumins>we started logging like 1hr ago
21:29:10  <pkrumins>and our test is like 1hr 15mins old
21:29:10  * wiwilliaquit (Ping timeout: 246 seconds)
21:29:18  <pkrumins>but i can clone your repo
21:29:30  <pkrumins>and do a silly push to get the hook running
21:29:34  <pkrumins>and see what happens
21:30:07  <slaskis>yeah sure
21:30:45  <slaskis>i'd expect at least chrome to pass as it passes in node mocha
21:31:44  <pkrumins>ok it's running
21:31:48  <pkrumins>looking at what's happening
21:32:01  <slaskis>fun!
21:32:01  <pkrumins>it's all syntax errors
21:32:06  <pkrumins>in all the browsers
21:32:10  <slaskis>no fun :(
21:33:17  <slaskis>i'm using those libs in browsers so i'm pretty sure it shouldn't have syntax errors
21:33:21  * _sorensenquit (Quit: Bye!)
21:35:05  * fotoveritejoined
21:40:20  * st_lukequit (Remote host closed the connection)
21:42:25  * juliangruber_joined
21:44:46  * shuaibquit (Quit: Textual IRC Client: http://www.textualapp.com/)
21:46:54  * juliangruber_quit (Ping timeout: 260 seconds)
21:49:19  * _sorensenjoined
21:56:57  <substack>hmmm
22:05:08  * fotoveritepart
22:05:17  * fotoveritejoined
22:12:26  * juliangruber_joined
22:22:16  * b__quit (Quit: Textual IRC Client: www.textualapp.com)
22:26:07  * st_lukejoined
22:26:21  * wiwilliajoined
22:30:06  * ralphtheninjajoined
22:32:20  * juliangruber_quit (Ping timeout: 248 seconds)
22:38:27  * dominictarrjoined
22:41:31  <dominictarr>rvagg, hey what up? are you gonna go to campjs.com in feb?
22:42:35  <rvagg>dominictarr: hadn't planned to
22:42:40  <rvagg>you're going to be in AUS then?
22:43:00  <dominictarr>I'll be in NZ, but close enough to make it maybe.
22:43:10  <dominictarr>THE ANTIPODES
22:43:39  <rvagg>I'll have to talk to Tim to see if he can tempt me
22:44:21  <Raynos>More PR! https://github.com/substack/tape/pull/11
22:47:03  <substack>Raynos: merged+published
22:47:14  <Raynos>Nice!
22:47:29  <Raynos>I should have also fixed those other two PRs
22:47:32  <Raynos>they should work now.
22:47:50  <substack>when tests pass I barely even care what the code is so long as there's not too much of it
22:49:41  <pkrumins>need more memory leak tests
22:51:37  <substack>true
22:51:50  <substack>http-raw has those
22:51:54  <substack>for the bouncy replacement
22:56:36  * pkruminstopic: http://browserling.com | http://testling.com | http://ci.testling.com | We put a browser in your browser!
22:58:59  * juliangruber_joined
22:59:34  <substack>created a new official {browser,test}ling help channel #browserling
22:59:43  <substack>this channel will remain unchanged
23:00:19  <substack>be as weird as you like
23:01:23  * st_lukequit (Remote host closed the connection)
23:03:50  <substack>need a more officialishious place to route all the businessy folk
23:05:13  * substacktopic: Unofficial browserling/testling mad science channel. For official help /join #browserling
23:05:27  <Raynos>substack: http://ci.testling.com/Colingo/valid-schema
23:05:31  <substack>yes
23:05:37  <Raynos>Fails in FF15. "object.keys" is not a function
23:05:50  <substack>looking
23:06:11  <substack>oh the ff instance might not've closed properly
23:06:11  <Raynos>btw that page only loads succesffully 1 / 10 times
23:06:25  <substack>loads every time here
23:06:26  <pkrumins>Raynos: loads 100% here always
23:06:46  <substack>aha
23:06:51  <pkrumins>yes that's the ff instance problem for sure.
23:06:57  <substack>ff wasn't closing so I closed it
23:06:57  <pkrumins>i'm working on that.
23:07:50  <substack>hmmm I wonder if this "no element found" issue is the race condition
23:07:56  <Raynos>substack: GET procect.js returns CSS for me
23:08:12  <Raynos>http://ci.testling.com/_/bundle/project.js
23:08:14  <Raynos>That is CSS
23:08:21  <pkrumins>it's js
23:08:23  <pkrumins>for me
23:08:24  <Raynos>its CSS for me
23:08:26  <substack>js here
23:08:30  <substack>Raynos: clear yo cache
23:08:34  <Raynos>oh now its js
23:08:40  <pkrumins>burn yo computer
23:08:56  <Raynos>xd
23:08:58  <substack>chrome's cache is so aggressive
23:09:08  <substack>frustratingly so
23:09:21  <Raynos>well I bumped the thing and FF test passed now
23:11:54  <Raynos>substack: http://ci.testling.com/Colingo/valid-schema.png can I make that IE8 orange
23:12:09  <Raynos>I want to tell my users that I dont support IE8 instead of the module is broken in IE8
23:12:43  <Raynos>so the flow would be if it passes its green, if it fails its orange, or warning, or grey or not supported or whatever. Just not the same as failure
23:13:47  <substack>just say that on the readme
23:14:35  <Raynos>substack: Also having ci.testling.com send me emails when it fails would be cool.
23:15:02  <Raynos>which would have to tie in with configuring the thing to be like "If this browser fails I dont care. Dont send me email"
23:17:21  * juliangruber_quit (Ping timeout: 265 seconds)
23:21:22  <pkrumins>we can do that
23:22:08  <dominictarr>substack, ci.testling.com needs to show the git commit messages. I don't keep a mental cache of the contents of commits indexed by their hash, unfortunately.
23:22:33  <substack>you can click a commit to show it on github
23:23:15  <pkrumins>although pretty neat idea
23:23:27  <pkrumins>as yeah you dont remember commits by hash
23:23:30  <pkrumins>not sure how hard that is
23:23:37  <pkrumins>oh should be pretty easy
23:23:44  <pkrumins>as the message comes together with the commit
23:23:50  <pkrumins>i'll look into it in 5 mins
23:24:19  <substack>but we need to store that in couch if we're not already doing that
23:24:53  * defunctzombie_zzchanged nick to defunctzombie
23:25:02  <pkrumins>oh
23:25:05  <pkrumins>but that's super easy isnt it
23:25:10  <pkrumins>just a string message
23:25:45  <substack>yes but the data needs to be threaded through several places in the app
23:25:50  <substack>and displayed in the ui
23:29:25  <dominictarr>substack, defunctzombie hey what happened with browserify 2?
23:30:09  <defunctzombie>dominictarr: not sure, I am still maintaining my version (https://github.com/shtylman/node-script) with some new ideas. substack and I have been talking about some of them
23:30:51  <defunctzombie>there is a lot of browserify code out there relying on some things that I think substack doesn't want to break
23:31:03  <defunctzombie>so that has to be handled more carefully
23:31:09  <dominictarr>what stuff?
23:32:08  <CoverSlide>like process.*
23:32:16  <defunctzombie>(require)(foo) style syntax is one, I think watch mode was another concern (not sure the status of that) and process.* yes
23:32:31  <defunctzombie>these days I am sad that process was not require('process')
23:32:34  <defunctzombie>but that ship has sailed
23:33:28  <substack>dominictarr: I've been busy with getting testling-ci lately
23:33:32  <substack>it's still in the queue
23:34:02  <substack>defunctzombie: and require('buffer') like it used to be
23:34:49  <defunctzombie>yea
23:34:57  <defunctzombie>that one is annoying as well
23:35:01  <dominictarr>what if we fixed all that in node2
23:35:07  <dominictarr>aka NO
23:35:09  <defunctzombie>node should not have introduced new globals
23:35:29  <defunctzombie>but it isn't that bad actually, cause you can detect them with esprima and such
23:35:38  <defunctzombie>and only add the browser shims if needed
23:35:46  <dominictarr>defunctzombie, we've been talking about refactoring out _all_ the code modules, and publishing them on NPM
23:36:14  <defunctzombie>I dunno
23:36:19  <defunctzombie>so something to think about is discovery
23:36:24  <defunctzombie>what is nice about the core docs
23:36:31  <dominictarr>okay. that is slightly more cleverness than is desirable
23:36:34  <defunctzombie>is that when I have to think about what http server to use.. I don't think
23:36:44  <defunctzombie>I just use what is "recommended"
23:36:56  <defunctzombie>and if that same type of discovery can be kept, then it is good
23:37:04  <defunctzombie>the old modules wiki was a good place because of that
23:37:17  <defunctzombie>you could just look through it for recommended modules to use
23:37:31  <defunctzombie>certainly there are things in core that should not be exposed
23:37:33  <defunctzombie>like utils
23:37:46  <defunctzombie>and other minor libs, but I think they are because core uses them
23:37:52  <defunctzombie>which makes other people automatically use them
23:39:05  <defunctzombie>I view npm as just a code delivery and discovery mechanism really
23:39:08  <dominictarr>well, what if there was a curation system
23:39:21  <dominictarr>... where you selected who you wanted to follow
23:39:37  <dominictarr>and then the modules you saw where recommended by those people
23:39:48  <dominictarr>but if you disagree with something, you just say that.
23:39:52  <defunctzombie>that could be interesting
23:40:00  <dominictarr>and then that disappears from your view
23:40:22  <defunctzombie>the thing is.. that sometimes it is more annoying to debate about a module than to just be recommended one to use
23:40:27  <defunctzombie>like take uuid module
23:40:28  <dominictarr>and you can also unfollow a user
23:40:33  <defunctzombie>(or node-uuid iirc)
23:40:40  <dominictarr>yes
23:40:43  <defunctzombie>I don't want to have to pick between 10 uuid modules
23:40:49  <defunctzombie>because 9 of them will be fucked up
23:40:50  <dominictarr>yes.
23:40:57  <dominictarr>I recommend node-uuid
23:41:01  <defunctzombie>so there are some things that are more like "system" level
23:41:03  <defunctzombie>yes
23:41:03  <dominictarr>so if you follow me, use that one
23:41:05  <defunctzombie>I think everyone does
23:41:08  <defunctzombie>interesting
23:41:15  <defunctzombie>that could be an interesting approach
23:41:19  <defunctzombie>developer curated
23:41:42  <dominictarr>I think I might have to experiment with this
23:41:45  <defunctzombie>for node-uuid what I actually tried to do was get that developer to get the uuid name
23:41:47  <defunctzombie>and he did
23:41:55  <defunctzombie>so eventually uuid will be the node-uuid
23:41:58  <defunctzombie>and that is good
23:42:05  <defunctzombie>these things can be resolved very well
23:42:11  <defunctzombie>npm is flexible
23:42:27  <dominictarr>I think there is already a uuid, that is different. it's a binding to a C lib
23:42:35  <defunctzombie>yea, but not for long
23:42:45  <defunctzombie>the node-uuid dev got added to it so he could replace it
23:42:50  <defunctzombie>the bindings are not maintained
23:43:00  <defunctzombie>and in this case it was good... the parties agreed
23:43:10  <defunctzombie>and we will get a more sane module name for the more commonly used module
23:43:21  <defunctzombie>it was a social problem that just required some mediation
23:44:08  <dominictarr>you could also allow users to tag modules
23:44:27  <dominictarr>but you'd ignore the tags from users you disagree with
23:44:44  <defunctzombie>hij1nx: component.json is not the right way forward imho, it is solving the wrong problem in a very confusing way
23:44:48  * juliangruber_joined
23:44:50  <defunctzombie>huh
23:46:22  * fotoveritequit (Quit: fotoverite)
23:46:29  <defunctzombie>dominictarr: the amusing thing about modules is that it is sort of a chicken and egg thing sometimes. If a module is popular, then I am more likely to use it (depending on what it does)
23:46:35  <defunctzombie>since I will benefit from crowd testing
23:46:56  <dominictarr>agreed
23:47:08  <dominictarr>but not everyone counts equally
23:48:21  <dominictarr>I'd take the recommendation of #stackvm residents over most others, for example
23:48:46  <defunctzombie>right
23:48:56  <defunctzombie>I do think the social recommendation aspects are interesting
23:49:14  <dominictarr>although, there are a few things that I might disagree with certain people about
23:49:29  <defunctzombie>honestly, I think it depends on the nature of the module
23:49:35  <dominictarr>so, I'd want to be able to express that, too.
23:49:51  <defunctzombie>there are things which really don't have "opinions"
23:49:56  <dominictarr>exactly
23:49:57  <defunctzombie>you are basically implementing a known algorithm
23:50:05  <defunctzombie>and it either works or it is broken
23:50:13  <defunctzombie>those are great things for general recommendation
23:50:20  <defunctzombie>but the higher level you go
23:50:24  <defunctzombie>the more flexible it gets
23:50:31  <dominictarr>yeah, if you are just implementing an algorithm
23:50:34  <chrisdickinson>it'd be kind of cool to do `npm search <terms> | npm vet`
23:50:35  <defunctzombie>like, I may not like the way core event emitters work
23:50:39  <dominictarr>you want the best implementation
23:50:40  <defunctzombie>but because I know everyone has them
23:50:45  <defunctzombie>and knows about them
23:50:49  <defunctzombie>I just stick to that
23:50:49  <dominictarr>but somethings like options parsing
23:50:53  <defunctzombie>right
23:50:58  <dominictarr>that is different
23:51:06  <dominictarr>I use optimist because it's popular
23:51:11  <chrisdickinson>open up the node repl with each of the results loaded and readmes accessible, and then when you quit out of the term you can write a comment in $EDITOR that applies to each of the modules, with a crosslink
23:51:11  <dominictarr>it's a standard
23:51:28  <defunctzombie>yea, and that is the info you want to be able to ask people about
23:51:38  <defunctzombie>I think node-toolbox was aiming to do similar stuff
23:51:42  <defunctzombie>but really just became a list iirc
23:52:13  <dominictarr>you get this problem with amazon and shit too
23:52:25  <chrisdickinson>actually that might be a really cool mini-review site: "<author> searched <xyz> and reviewed <these modules>:"
23:52:30  <dominictarr>where some peolpe love books about scientology or creationism or whatever
23:52:31  <chrisdickinson>and then you could follow them, etc.
23:52:35  <dominictarr>but other people hate them
23:52:58  <dominictarr>or evolution
23:53:20  <dominictarr>so those books get lots of 5 stars, and lots of 1 stars
23:53:36  <chrisdickinson>thoughts on that? maybe even bringing in previous reviews of those modules to the repl somehow so you've got everything you need to make a decision directly from a search command?
23:53:36  <dominictarr>the number of stars isn't really useful
23:53:55  <dominictarr>what is important in whether or not you'd give that book a good or bad review
23:54:33  <chrisdickinson>if the text "i used / i ended up using <x>" appeared in the review, you could even push that into the review title
23:54:34  <dominictarr>well, I've been thinking about this a while actually
23:54:52  <chrisdickinson>"garybusey searched for uuid and went with node-uuid (like N others)"
23:55:21  <dominictarr>well, I don't necessarily care what garybusey used
23:55:33  <chrisdickinson>if you're following him, though.
23:55:57  <dominictarr>if you agree with him, then yes.
23:56:52  <chrisdickinson>so the flow would be, do a search, grab results, grab reviews that match that set of modules for those terms, display overall stats per module, and if you're following anybody that reviewed those modules, say which one they went with.
23:56:59  <chrisdickinson>and drop you into a repl to play with each of the modules.
23:57:07  <Raynos>dominictarr: I experimented with the npm following thing in npm-used
23:57:19  <chrisdickinson>maybe with a command to bring up reviews, like ".why garybusey"
23:57:41  <dominictarr>I think i'd want a web interface primarily
23:57:48  <Raynos>I think the community curation
23:57:50  <Raynos>is optimum
23:57:54  <dominictarr>but I guess you could do either
23:57:59  <Raynos>It should be based on real usage
23:58:01  <Raynos>not opinions
23:58:10  <Raynos>it should also be weighted to now
23:58:12  <Raynos>and not 3 months ago
23:58:13  <chrisdickinson>yeah, for a lot of it a web interface makes more sense, but you really want to be able to try out each of the modules from your search
23:58:19  <Raynos>Actually the weighting algorithm
23:58:22  <Raynos>should be customizable
23:58:23  <dominictarr>you could have an opt in hook that tracked you actual calls to npm install
23:58:43  <Raynos>dominictarr: or spider all my modules and the dependencies listed in package.json
23:59:02  <dominictarr>well, spider your dev folder
23:59:11  <dominictarr>but don't necessarily drill down
23:59:29  <chrisdickinson>Raynos: that'd be interesting, a profile site that showed the gestalt of the node_modules tree across an author's history of repos
23:59:29  <Raynos>spider npm
23:59:32  <dominictarr>because installing express was your decision
23:59:32  <Raynos>I have 170 on npm
23:59:42  <dominictarr>but the sub deps where tj's