00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:00:25  <dominictarr>gozala: that is the same thing as not testing
00:00:26  <Raynos>dominictarr: createServer(function (stream) { stream.pipe(stream) }) like an echo server
00:00:46  <dominictarr>if you are gonna do that, you should test you test framework with a different test framework
00:00:48  <gozala>dominictarr: I disagree
00:00:54  <gozala>every software is broken
00:01:07  <gozala>if you test software with any other software
00:01:15  <gozala>it's broken too
00:01:23  <gozala>but that is not the same as not testing at all
00:01:25  <dominictarr>Raynos: so, I think you should call it "" (empty string)
00:01:37  <dominictarr>because it doesn't do anything
00:01:51  <dominictarr>well, it's just a plain through stream.
00:02:13  <Raynos>but its a duplex passthrough stream
00:02:22  <Raynos>I guess thats the word
00:02:49  <dominictarr>it's a through stream because it has no side-effects
00:03:04  <dominictarr>it's sideeffects that make a duplex not a through.
00:03:51  <dominictarr>gozala: yes, but if you use a different test framework, it will have different bugs to your test framework
00:05:27  * yorickquit (Remote host closed the connection)
00:22:00  <gozala>anyway to find out if my tests actually runnig https://github.com/gozala/retape ?
00:26:29  <Raynos>prob not
00:26:31  <Raynos>http://ci.testling.com/gozala/retape does nothing
00:29:09  <gozala>Raynos: Yeah problem was lower case gozala
00:32:58  * dominictarrquit (Ping timeout: 245 seconds)
00:51:16  <mbalho>GOZALA
00:57:15  * jibayjoined
01:04:11  <substack>hooray http://ci.testling.com/Gozala/retape
01:04:14  <substack>more tests!
01:04:20  <substack>testing: pretty great
01:06:14  <gozala>substack: it's just a workaround for me not to rewrite all tests in tape :)
01:08:11  * defunctzombie_zzchanged nick to defunctzombie
01:09:00  <substack>all testling cares about is that TAP-formatted output is getting console.logged
01:54:19  <mbalho>substack: i just pushed a version of blocking that kind of works, you can still fall through walls
01:58:11  <defunctzombie>substack: what do you do about 'process.on('uncaughtException')' on the browser?
02:02:32  <defunctzombie>gozala: I like that commonjs test stuff
02:02:44  <defunctzombie>gozala: test(name, function(assert, done)) feels quite clean
02:03:04  <Raynos>GAH.
02:03:27  <Raynos>defunctzombie: that's the same as tape
02:03:33  <Raynos>except tape has assert.end() instead of done()
02:03:35  <defunctzombie>Raynos: almost
02:03:44  <defunctzombie>yea, don't like that as much
02:03:46  <defunctzombie>for one reason
02:03:51  <defunctzombie>sync tests
02:04:25  <defunctzombie>let me gist the api style I would like more and see what you think
02:04:56  <Raynos>I know what you mean
02:06:03  <defunctzombie>Raynos: https://gist.github.com/4440171
02:06:55  <defunctzombie>after trying all sorts of different test runners, harnesses, etc, I find that to be the one with the smallest overhead (I think)
02:07:03  <defunctzombie>ideally you could use the assert module
02:07:07  <defunctzombie>but the problem is async tests
02:07:15  <defunctzombie>so it becomes cleaner to pass it in as an argument
02:07:48  <Raynos>wth is that extra
02:07:52  <Raynos>OUT OF SCOPE
02:08:06  <defunctzombie>?
02:09:35  <Raynos>but I agree
02:09:41  <Raynos>that's ideal
02:10:23  <Raynos>defunctzombie: https://gist.github.com/4440178#file-gistfile1-js-L15
02:10:56  <defunctzombie>sure, just means I have to do that every time I start a new test file
02:11:24  <defunctzombie>but yea, same thing
02:11:37  <defunctzombie>I am actually thinking that 'test' should go away
02:11:52  <defunctzombie>and just module.exports['foo'] = function(assert, done);
02:12:06  <defunctzombie>or exports['foo'] = function(assert, done)
02:12:09  <Raynos>https://gist.github.com/4440178
02:12:11  <defunctzombie>cause then you can use anny runner
02:12:25  <Raynos>module.exports['foo'] is bad
02:12:31  <Raynos>because then you cant use node as your runner
02:12:44  <Raynos>its also a pain in the ass to read
02:12:56  <defunctzombie>you can use node as the runner
02:13:03  <defunctzombie>by just having a test.js file
02:13:05  <Raynos>not without boilerplate
02:13:21  <Raynos>I prefer not to export my tests, its ugly
02:13:27  <Raynos>you can already use any runner you want
02:13:30  <Raynos>its called output TAP
02:13:32  <Raynos>on stdout
02:13:33  <defunctzombie>I guess the thought is that the runner is not anything special
02:14:04  <defunctzombie>it just reads modules and runs stuff from them
02:14:12  <defunctzombie>and if it passes in a tap like assert then great
02:14:26  <defunctzombie>or if it passes in some other assert (for local browser testing and nice display) then that is cool too
02:14:37  <defunctzombie>basically, the runner decides what output it wants
02:14:56  <Raynos>but then it doesnt work without a runner
02:15:39  <defunctzombie>unless you have a runner js file
02:15:58  <defunctzombie>(which I am inclined to use more often now since it makes tests run in specific order)
02:16:17  <defunctzombie>but yes.. by default a test file is nothing more than a module
02:17:08  <Raynos>i do have a runner js file
02:17:23  <Raynos>https://github.com/Raynos/push-queue/blob/master/test/index.js
02:17:27  <Raynos>it has a bunch of require statements
02:17:32  <defunctzombie>personally I don't like the idea that tap consumes my output in node. How the test runner wants to gather info is none of my business
02:17:47  <defunctzombie>yea, you could do that.. or provide a nice globber or something
02:17:51  <defunctzombie>for the generic case
02:17:56  <Raynos>thats a pain
02:17:59  <defunctzombie>most people end up using a runner anyhow
02:18:00  <Raynos>because of browserify support
02:18:25  <Raynos>index.js is the simplest way to do it with browserify
02:18:33  <Raynos>the alternative is auto generated index.js with a list of all requires
02:19:11  <defunctzombie>I guess the question is what is the goal for browserifing the tests
02:19:38  <defunctzombie>so something like testling could support it through the files: */*.js field
02:19:46  <defunctzombie>and use whatever runner it wants
02:19:56  <defunctzombie>local testing still needs some sort of runner
02:20:15  <defunctzombie>and that thing would just support loading files or whatnot
02:20:18  <Raynos>defunctzombie: */*.js is a disaster.
02:20:25  <Raynos>half the files are util files
02:20:29  <defunctzombie>just an example
02:20:36  <Raynos>and im not going to add -test.js to the end of test files
02:20:36  <substack>browserify test/*.js -o bundle.js; echo '<script src="bundle.js"></script>' > index.html; xdg-open index.html
02:20:41  <substack>is testling locally
02:20:52  <substack>just look at the console.logs in the debugger
02:20:55  <defunctzombie>substack: that is a terrible way to test locally cause it doesn't give good feedback
02:21:04  <defunctzombie>substack: that is why I made zuul for my mocha projects
02:21:06  <Raynos>im just going to enumerate them in index.js
02:21:17  <Raynos>defunctzombie: have you tried using testem?
02:21:25  <substack>I'll have something fancier eventually
02:21:28  <defunctzombie>Raynos: seemed too comeplex
02:21:31  <defunctzombie>*complex
02:21:36  <Raynos>yeah it is a beast
02:21:38  <defunctzombie>I don't want to set more shit up.. ever
02:21:42  <defunctzombie>zuul you just run
02:21:42  <Raynos>:D
02:21:43  <defunctzombie>and it works
02:21:48  <substack>defunctzombie: window.onerror
02:21:49  <Raynos>true
02:21:54  <Raynos>the testem config is a pain in the ass
02:21:57  <substack>is the uncaughtException analogue
02:22:03  <defunctzombie>yea.. that is why I don't use it
02:22:04  <defunctzombie>substack: gotcha
02:22:17  <defunctzombie>really the reason I think the exports stuff is interesting
02:22:23  <defunctzombie>is because it disconnects the tests
02:22:30  <defunctzombie>from whatever runner/output you want
02:22:37  <defunctzombie>since the two do not have to be connected
02:22:58  <defunctzombie>and provides for a bit more flexibility on the tooling side as a result
02:23:11  <substack>that is true
02:23:14  <defunctzombie>this is what mocha leverages for the cool web display
02:23:17  <substack>it's more annoying to run the tests though
02:23:23  <substack>you need to use a harness program
02:23:27  <defunctzombie>yea
02:23:34  <substack>with the other approach you can just run the files directly
02:23:37  <defunctzombie>so just thinking about if that is really a downside
02:23:39  <substack>trade-off
02:23:49  <defunctzombie>as most people use some harness or runner anyhow
02:24:03  <substack>I run tests directly without a harness all the time
02:24:10  <substack>I only use the harness when I want to run all the tests in the suite
02:24:38  <defunctzombie>i usually just type 'mocha' or 'mocha --grep' or something of the sort
02:24:43  <substack>it's really nice when you're writing a test in the first place since you can just build up the assertions incrementally from an example file
02:24:44  <defunctzombie>to run anything I need
02:24:54  <defunctzombie>agreed
02:25:21  <defunctzombie>I guess I view it as more of.. "here are my tests" consume them as you wish
02:25:35  <Raynos>...
02:25:46  <Raynos>"here are my tests. It's a process you spawn. the process outputs TAP" consume them as you wish
02:25:55  <defunctzombie>don't like that ;)
02:26:00  <defunctzombie>not as flexible
02:26:03  <Raynos>thats how you do shit
02:26:03  <substack>I like that a lot
02:26:07  <defunctzombie>and annoying to instrument around
02:26:08  <Raynos>its the most flexible thing ever
02:26:11  <Raynos>its unix
02:26:14  <Raynos>its a command
02:26:22  <Raynos>pipe it around
02:26:39  <defunctzombie>I don't like things messing with my stdout haha
02:26:41  <Raynos>and its isolated at the process level
02:26:42  <substack>THIS IS UNIX. I KNOW THIS.
02:26:47  <Raynos>which is win
02:26:50  <defunctzombie>when I console something I want to see it
02:26:58  <defunctzombie>hahaha
02:27:03  <defunctzombie>oh man.. I should watch that movie
02:27:17  <defunctzombie>tap output is great... until it isn't
02:27:19  <substack>defunctzombie: you can console.log() whatever you want in tap tests
02:27:21  <defunctzombie>this is not tied to any output
02:27:37  <substack>tap only cares about lines that start with /^ok/, /^not ok/, and /^1..\d/
02:28:16  <substack>I especially like how I can have tap and tape tests in the same repo
02:28:16  <Raynos>defunctzombie: stderr exists for a reason
02:28:24  <substack>since they both are just programs that output tap output
02:28:37  <substack>in fact for my tape tests in node I still use the `tap` command as the harness to run the whole suite
02:28:40  <Raynos>but I do agree
02:28:42  <Raynos>raw tap output is annoying
02:28:47  <Raynos>I want a tool to parse and aggregate it
02:28:47  <defunctzombie>I started to look at what it would take to make pretty displays for tape output in browser
02:29:13  <defunctzombie>tap output is great if you want tap output
02:29:19  <defunctzombie>but why do I care about tap output?
02:29:33  <defunctzombie>my runner tool (in the case of testling or whatnot) could easily hoist assert
02:29:39  <defunctzombie>and grab whatever it wants
02:29:40  <substack>anyways, TIME TO FINISH browserify.org already
02:29:43  <defunctzombie>without any output
02:29:51  <defunctzombie>jesus.. have you not finished that yeat?
02:29:52  <defunctzombie>*yet?
02:30:00  <substack>auth remains
02:30:06  <defunctzombie>you need to finish it so my browser range change can land!
02:30:08  <substack>well I wrote a ton of tiny libs
02:30:22  <substack>since it's something of a showcase for doing modular development
02:31:12  <defunctzombie>Raynos: for the case where you don't want to write a runner.js by hand, you could just have the runner program support generating one for you or something
02:31:34  <defunctzombie>and in the cases where you don't care, you just put the runner in scripts of package.json as most do
02:33:13  <defunctzombie>Raynos: this is what I was looking at https://github.com/Gozala/test-commonjs
02:33:37  <defunctzombie>except without all the baggage
02:33:42  <defunctzombie>just the simple spec
02:34:52  <Raynos>defunctzombie: I implemented commonJS 2 years ago
02:34:56  <Raynos>that is a disaster
02:35:04  <Raynos>tape needs a little api sugar
02:35:14  <defunctzombie>I wouldn't do the whole spec stuff they have there
02:35:24  <defunctzombie>just the simple api I gisted
02:35:33  <defunctzombie>to define how a test function looks, that is it
02:35:44  <Raynos>tape already does most of it
02:35:48  <Raynos>ill just wrap tape with api sugar
02:35:49  <defunctzombie>yes
02:35:53  <Raynos>Ugh I cant -.-
02:35:56  <Raynos>I have to patch tape
02:36:05  <defunctzombie>I agree, I just don't love tap like tall do
02:36:16  <defunctzombie>I think it is great for when you want tap, but not as useful in other cases haha
02:36:23  <Raynos>you dont get it
02:36:28  <Raynos>you want any arbitary reporter?
02:36:36  <Raynos>Sure pick any of these 20 reporters just stream tap into them
02:36:54  <defunctzombie>sure.. you are just describing standards
02:36:56  <defunctzombie>:)
02:36:57  <Raynos>childprocess.stdout.pipe(reporter("whatever the fuck")).pipe(process.stdout)
02:37:05  <defunctzombie>anytime you have a standard people make common tools
02:37:11  <defunctzombie>the standard could be anything tho
02:37:19  <Raynos>the standard is TAP
02:37:31  <defunctzombie>but why are you telling me my tests have to output tap all the time forever
02:37:40  <substack>tap has been around since the 80s
02:37:40  <defunctzombie>easier if the runner just decides that :)
02:37:50  <defunctzombie>doesn't make it good :)
02:37:54  <defunctzombie>just makes it around since the 80s
02:38:09  <substack>it has staying power
02:38:15  <defunctzombie>(I am trolling, but really the point is that the tests != the output != runner)
02:38:39  <substack>UNIX.
02:38:55  <defunctzombie>If I want to output my test shit in json I can easily do that if I did not make assumptions about output during writing my tests
02:38:59  <hij1nx>tap +1
02:39:12  <substack>you can still output json
02:39:14  <substack>tap doesn't care
02:39:18  * ralphtheninjaquit (Ping timeout: 244 seconds)
02:39:36  <Raynos>defunctzombie: you misunderstand what a test is
02:39:38  <hij1nx>test anything p[rotocol|operly]
02:39:39  <Raynos>a test is a process
02:39:40  <Raynos>that outputs TAP
02:39:53  <defunctzombie>not really
02:40:00  <Raynos>well really
02:40:04  <defunctzombie>a test is a program that exists with non zero when it fails
02:40:08  <Raynos>it's a process that has a zero or non zero exit code
02:40:12  <Raynos>but you use TAP to be nice
02:40:12  <defunctzombie>in all reality :)
02:40:15  <defunctzombie>look at travis ci
02:40:16  <substack>browsers don't have exit codes
02:40:18  <Raynos>and represent actual meta data about errors
02:40:18  <defunctzombie>they run tests all the time
02:40:20  <defunctzombie>no tap
02:40:22  <defunctzombie>no problem
02:41:06  <defunctzombie>no reason the fact that my tests can output tap has to be baked into them
02:41:09  <hij1nx>substack: can you cite some proof of that?
02:41:35  <defunctzombie>mocha is a clear example of this
02:41:37  <hij1nx>substack: it just sounds odd.
02:41:38  <substack>hij1nx: they daemonize themselves and don't exit when their event loop is empty
02:41:45  <Raynos>defunctzombie: mocha is a troll :D
02:41:54  <defunctzombie>sure.. but it is still an example
02:41:57  <Raynos>yeah sure
02:42:00  <Raynos>use a testing framework
02:42:02  <defunctzombie>I didn't have to re-write a single test
02:42:05  <Raynos>have the testing framework output whatever who cares
02:42:07  <defunctzombie>and it works
02:42:15  <defunctzombie>yes.. that is my point!
02:42:20  <defunctzombie>the test author DOES NOT CARE
02:42:21  <hij1nx>substack: ok that makes sense.
02:42:35  <Raynos>UNLESS
02:42:36  <substack>so long as your tests can output tap so that testling-ci can parse the results
02:42:38  <mbalho>guys i dont know what the argument is about but i just wanted to pop in and say you are both wrong
02:42:39  <defunctzombie>you telling me my tests need to output TAP means NOTHING to me
02:42:39  <Raynos>said test author wants to run his tests
02:42:43  <Raynos>in many different environments
02:42:57  <Raynos>on many different tools
02:42:59  <Raynos>and many different platforms
02:43:11  <defunctzombie>then use a runner that does that
02:43:15  <Raynos>in which case you play nice and use a protocol
02:43:16  <substack>speaking of, I need to write a tap parser because tap/lib/consumer.js is giving me problems :(
02:43:23  <defunctzombie>it will most likely be different on each platform anyway
02:43:36  <Raynos>meh
02:43:38  <defunctzombie>substack: see if you didn't care about tap you would not need a tap parser :p
02:43:39  <Raynos>bikesheddery
02:43:44  <Raynos>and useless conversations
02:43:44  <defunctzombie>you could just pass teh assert you wanted
02:43:52  <defunctzombie>:p
02:44:02  <Raynos>...
02:44:04  <defunctzombie>how many times have I told you all programming is bikesheddery
02:44:07  <substack>defunctzombie: tried that. didn't work
02:44:11  <substack>see testling.com
02:44:23  <defunctzombie>substack: why didn't it work?
02:44:42  <substack>people are very opinionated when it comes to testing
02:44:53  <defunctzombie>yes
02:45:53  <Raynos>defunctzombie: https://github.com/defunctzombie 404
02:46:10  <defunctzombie>?
02:46:31  <defunctzombie>Raynos: you lost me
02:46:31  <Raynos>i dont have a mental mapping between github <-> irc aliases
02:46:37  <defunctzombie>hahaha
02:46:54  * defunctzombieshtylman
02:47:11  <Raynos>:D
02:51:36  * jibayquit (Quit: Leaving)
02:53:12  * jez0990_quit (Ping timeout: 276 seconds)
03:15:33  <mbalho>oo https://github.com/thesmart/node-SmartStream
03:25:46  * tphummelquit (Quit: tphummel)
03:51:45  * mikealquit (Quit: Leaving.)
03:52:15  * mikealjoined
04:01:29  <mbalho>substack: http://www.engadget.com/2005/03/30/underdog-stinky-robot-beats-mit-in-underwater-bot/
04:04:38  <substack>looks like the MATE contest but I don't see a mention of it
04:09:14  * notelliottcable_changed nick to notelliottcable
04:15:13  <notelliottcable>substack: I can't remember, did I alert you of the pull request I left on optimist?
04:15:27  <substack>I saw it
04:15:30  <substack>haven't gotten to it yet
04:15:41  * notelliottcablenods
04:16:01  <notelliottcable>it's not remotely a blocker for me, just thought I'd leave an issue for it
04:16:03  * substackhas updated 21 npm packages this month
04:16:11  <substack>yep, appreciated!
04:16:17  <notelliottcable>bro's busy, ain't'ee
04:16:28  <notelliottcable>side-note: I remember I introduced you to Paws, at one point.
04:16:43  <substack>also check this out: http://browserling.com:9009/
04:16:45  <notelliottcable>the PoC implementation I was writing at the time, has advanced to the point of a working REPL
04:16:48  <mbalho>is that a cat that you own, if yes i need photos
04:16:51  <substack>especially http://browserling.com:9009/search
04:17:33  <notelliottcable>https://vimeo.com/56291617
04:17:57  <mbalho>aww
04:18:11  <notelliottcable>(the video shows a serious speed issue that turned out to be due to a stupid mistake on my part that I killed off within the next 24 hours; it's now blazing fast, as it should have been)
04:18:19  <notelliottcable>here's the REPL: http://ell.io/iA1Ze
04:19:59  <notelliottcable>unfortunately, attempting to write completely unabstracted code in Paws is a pain-in-the-ass (unsurprisingly),
04:20:20  <notelliottcable>ex. https://github.com/elliottcable/muPaws.js/blob/7658de6/examples/09.a.routine.paws
04:21:19  <notelliottcable>and, to boot, attempts to write even more complex examples have uncovered some serious design flaws … but that, at least, is *good* news, as the entire point of a proof-of-concept like this is to allow us to uncover conceptual flaws in something that's grown too large to hold in my head all-at-once.
04:21:26  <notelliottcable>just some updates for you. (=
04:21:57  <notelliottcable>substack: I love the header, did you draw that?
04:23:45  <notelliottcable>substack: also, to be clear … it's not mentioned on the page, but is testling-ci *only* for browserling modules? as what I'm trying to build is a direct competitor to browserling, soooooo … it obviously won't, y'know, work *inside* browserling
04:24:59  <substack>browserify you mean?
04:25:06  <notelliottcable>yeah, sorry :3
04:25:13  <substack>should work for that
04:25:28  <substack>you can use the "scripts" field instead of "files"
04:25:47  <notelliottcable>not sure of the details, haven't looked into either right now. Will come play later.
04:26:04  <substack>"files" are run through browserify but "scripts" just plop down a <script> tag directly on the page
04:26:12  <notelliottcable>love you, and generally love your products, but had a *lot* of issues with Testling last time around. It just wasn't appropriate for what I was trying to do, or something, I suppose.
04:26:26  <notelliottcable>Will give it another shot now that I'm returning to the project after a few months, though. Looks like you've made some real progress.
04:26:32  <substack>testling-ci is turning out to be a lot more robust
04:26:42  <notelliottcable>testling-ci sounds like a *huge* step up from the, honestly, rather arcane and painful process for using testling previously
04:26:54  * tphummeljoined
04:27:09  <notelliottcable>just throwing this out there: an excellent next-step for you, as an avid user who *genuinely* needs your products … finding a way to improve upon error-reporting.
04:27:18  <notelliottcable>or, *most*-ideally, find a way to integrate with an equivalent of node-inspector;
04:27:37  <notelliottcable>something to pipe debugging controls from any browser you can possibly obtain them in, to a central debugger attached to testing.
04:27:38  <substack>I have some projects to do instrumentation
04:27:39  <substack>in the works
04:27:40  <notelliottcable>testling*
04:27:43  <notelliottcable>excellent!
04:28:03  <substack>we'll have some integration with browserling for live debugging too
04:28:16  <notelliottcable>side-note: if you've ever got something you need particular help on, and it's disjoint enough from the rest of your ecosystem to not need me to go learn twelve dozen of your modules in-depth to be able to help … come poke me, anytime. You've always got interesting goals.
04:28:48  <notelliottcable>though I'm likely to be busy in the immediate future. moving to Chicago.
04:29:15  <notelliottcable>wait, th'fuck? you went to UAF!? How'd you end up in AK?
04:31:42  <substack>my parents live there
04:32:03  <substack>jesusabdullah went to uaf too
04:32:30  <notelliottcable>Did you grow up in AK?
04:32:51  <notelliottcable>Anchorage, here. Attended UAA, but only for a semester or two. Moved to Montana.
04:33:16  <substack>my parents moved to alaska when I was in high school
04:52:30  <notelliottcable>substack: what else are you planning to do with the testling-ci PROFESH EDITION, other than private repos?
04:52:48  <substack>just private repos
04:54:43  <substack>unless you can think of things that would be fitting for the PROFESH edition aside from that
04:57:06  <substack>I'll also be doing more work on schoolbus to get it to play nicely with the undocumented "server" field
04:57:26  <substack>to start attacking the usual selenium use-case
05:00:43  * tphummelquit (Quit: tphummel)
05:01:56  <notelliottcable>no idea what either of those mean.
05:02:07  <notelliottcable>is testing non-ci going to “stay a thing?”
05:03:01  <notelliottcable>perhaps unlimited time on the cluster, with a user-side “guard”/watch style re-run the tests every time you save thing
05:03:08  <notelliottcable>CI-on-save instead of CI-on-commit
05:19:03  <mbalho>lolz http://sudoroom.org/ai1ec_event/wikitown-maker-village-meetup/?instance_id
05:20:39  <substack>wikiville
05:21:11  <substack>glog auth nearly works
05:21:56  <mbalho>my mom just sent me this video http://www.youtube.com/watch?v=KD9qvF4z8cE
05:23:01  * tphummeljoined
05:38:02  * tphummelquit (Quit: tphummel)
05:41:16  * tphummeljoined
05:55:09  * tphummelquit (Quit: tphummel)
06:20:34  * dguttmanquit (Quit: dguttman)
06:23:42  * tphummeljoined
06:23:47  * tphummelquit (Client Quit)
06:30:02  <defunctzombie>substack: testling ci needs a favicon :)
06:42:51  <defunctzombie>if a testling-ci browser is not showing up, does that mean it has stalled?
06:44:38  <st_luke>ie9?
06:45:21  <substack>defunctzombie: yeah, the means the browser never gave us any info back
06:45:31  <defunctzombie>yes, ie9
06:45:32  <defunctzombie>haha
06:45:38  <substack>https://github.com/substack/glog#authenticating-glog
06:45:42  <substack>just published this!
06:49:35  <st_luke>awesome
06:53:03  <defunctzombie>substack: I have firefox 15 and all my tests pass onit
06:53:09  <defunctzombie>however on testling one fails
06:53:10  <defunctzombie>:(
06:53:23  <defunctzombie>the error is not so helpful to say the least
06:55:20  <defunctzombie>also.. for browserling.. do you clear the clipboard hahaha
06:55:38  <defunctzombie>all tests pass on browserling version of FF 15
06:59:28  <gozala>substack: it looks like github caches images from readme
06:59:38  <gozala>so updates for test don't show up
06:59:38  <gozala>https://github.com/Gozala/interset
06:59:53  <gozala>on readme you can see two browser logos
07:00:06  <gozala>but if you click there are bunch of them
07:00:22  <defunctzombie>gozala: I see a bunch of logos
07:00:53  <gozala>maybe they cache per user session or something I don't know
07:01:24  <gozala>but when I look at gh page source img src is https://a248.e.akamai.net/camo.github.com/e1e0a920e49f2c8d18a6536bd094cbd8f466b9be/687474703a2f2f63692e746573746c696e672e636f6d2f476f7a616c612f696e7465727365742e706e67
07:01:51  <gozala>instead of actual http://ci.testling.com/Gozala/interset.png
07:02:16  <gozala>defunctzombie: I suspect you also get some github url there instead of the ci.testling.com
07:02:23  <defunctzombie>yes
07:02:41  <gozala>presumably that's because of https / http
07:03:10  <gozala>including no https content on http pages are considered insecure
07:05:13  <gozala>substack: defunctzombie travis.ci works fine because they use https urls
07:05:38  <defunctzombie>sounds like someone needs to get a cert
07:35:01  <substack>holy crap al jazeera bought current tv
07:47:24  <jesusabdullah>who is that
07:48:55  * lyaunzbequit (Ping timeout: 260 seconds)
08:20:52  <substack>jesusabdullah: it's you!!
08:23:08  <jesusabdullah>it is!
08:31:45  * AvianFluquit (Remote host closed the connection)
08:33:09  * jdenchanged nick to jden|zz
08:39:19  * No9joined
08:40:38  * defunctzombiechanged nick to defunctzombie_zz
09:07:37  * devaholicjoined
09:35:05  <jesusabdullah>substack: cure lil' blog engine you got there
09:36:36  <jesusabdullah>cute*
09:39:16  <substack>auth is a bit of a hack
09:39:24  <substack>it's a second github repo with a users.json in it
09:39:24  <jesusabdullah>so's life
09:39:28  <substack>s/github/git/
09:39:36  <jesusabdullah>nice
09:39:46  <jesusabdullah>that's basically how gitolite manages users
09:42:52  <jesusabdullah>how does publish date work?
09:43:01  <jesusabdullah>do you just extract the date of first commit?
09:43:05  <substack>tags
09:43:11  <jesusabdullah>I see
09:43:23  <substack>glog publish filename.markdown 'title text'
09:43:27  <substack>makes an annotated tag
09:43:52  <substack>that way you can track half-finished articles in the same git repo
09:45:55  <jesusabdullah>word
09:46:27  <jesusabdullah>I'm still using jekyll
09:46:42  <jesusabdullah>haven't had a compelling reason to go through the trouble of changing it up
09:46:48  <jesusabdullah>it's hosted by github after all
10:04:32  <jesusabdullah>oh substack does your thinger support .md files as well? I like to ue .md not .markdown
10:04:36  <jesusabdullah>due to length
10:23:36  <substack>I think so
10:25:31  <jesusabdullah>I'm gonna hold you accountable!
10:25:35  <jesusabdullah>You just wait and see!
10:25:48  <jesusabdullah>God the internet here sucks so many balls
11:08:08  <st_luke>I love bots that test PRs against the test suite
11:08:22  <st_luke>especially when someone has a broken build and your PR fixes it
11:09:38  * jibayjoined
11:12:58  * dominictarrjoined
11:23:47  <st_luke>wow 6am
11:28:35  * ralphtheninjajoined
11:33:52  <dominictarr>ralphtheninja: hey whats up?
11:36:38  * dominictarr_joined
11:37:12  * st_lukequit (Remote host closed the connection)
11:37:54  <dominictarr_>TEST
11:38:03  * dominictarrquit (Quit: Page closed)
11:38:04  * dominictarr_changed nick to dominictarr
12:24:32  * ralphtheninjaquit (Ping timeout: 255 seconds)
13:13:57  * yorickjoined
13:13:57  * yorickquit (Changing host)
13:13:57  * yorickjoined
14:05:03  <substack>http://substack.net/images/oscilloscope.png
14:13:07  <dominictarr>oh, this is for baudio
14:13:11  <dominictarr>we need this
14:13:17  <dominictarr>also, FFT
14:54:44  <substack>http://substack.net/images/tape_drive.png
15:36:55  * dominictarrquit (Quit: dominictarr)
15:39:28  * AvianFlujoined
15:43:42  * jez0990_joined
15:57:02  * AndChat648704quit (Ping timeout: 252 seconds)
15:58:34  * jez0990_quit (Ping timeout: 244 seconds)
16:03:37  * AndChat648704joined
16:07:11  * dguttmanjoined
16:55:01  * mikealquit (Quit: Leaving.)
16:58:16  * defunctzombie_zzchanged nick to defunctzombie
17:09:12  <defunctzombie>substack: what do you draw with?
17:17:43  * mikealjoined
17:26:55  * lyaunzbejoined
17:27:54  * dguttmanquit (Quit: dguttman)
17:47:39  <ehd>need project_questor.png
18:20:05  <notelliottcable>substack: ditto defunctzombie
18:30:03  * dguttmanjoined
18:47:30  * Dspriggsjoined
19:00:32  * dguttmanquit (Quit: dguttman)
19:01:41  * st_lukejoined
19:04:50  * dguttmanjoined
19:11:49  <chrisdickinson>out of curiosity: how does #stackvm feel about the promises spec?
19:12:17  <defunctzombie>meh
19:12:22  <chrisdickinson>(i've never felt the need to use them, personally -- i mostly stick to the node-style callback / event emitter / stream hierarchy)
19:20:55  <mbalho>when i have a callback and needs to fire after N callbacks have fired i use an external lib
19:20:58  <mbalho>but not a promise lib
19:21:07  <mbalho>i like a simple queue
19:21:15  <mbalho>async.parallel or async.series etc
19:41:31  * chrisdickinsonnods
19:43:59  * dguttmanquit (Quit: dguttman)
19:45:59  * tilgovijoined
19:59:00  * dguttmanjoined
20:03:37  * tilgoviquit (Ping timeout: 265 seconds)
20:41:46  <CoverSlide>i usually roll my own.
20:41:51  <CoverSlide>have a counter
20:42:01  <CoverSlide>and a "done" callback
20:43:29  * ITprojoined
20:43:43  * dguttmanquit (Quit: dguttman)
20:45:10  <defunctzombie>mbalho: CoverSlide: https://github.com/Raynos/after
20:45:47  * dguttmanjoined
20:47:35  <mbalho>ooh
20:47:46  <CoverSlide>interesting
20:48:14  <defunctzombie>question.. lets say I have some external api I expose to users. I also want to make use of the api by my website as well (since there might be overlap) what do you guys do in this case?
20:48:25  <defunctzombie>(assume the external api requires api keys or such)
20:48:32  <defunctzombie>do you create an internal only copy?
20:48:41  <defunctzombie>or do you make a super key for your site to use?
20:49:02  <defunctzombie>pros/cons of the various approaches?
20:49:14  <defunctzombie>so far I have generally had copies and kept the site used one only for the site
20:49:34  <mbalho>1 api = less work
20:49:55  <defunctzombie>I agree
20:50:11  <defunctzombie>so what do you do about website access to it?
20:50:51  <defunctzombie>do you go with the oauth type approach and create keys for the site to use?
20:50:54  <defunctzombie>or what
20:51:58  <mbalho>yea exactly
20:52:26  <defunctzombie>hm
20:52:45  <defunctzombie>you just bypass the user confirming it I guess?
20:53:07  <mbalho>in gather we hardcode this in the front end "/login?client_id=fd51ccc4875d66076f94b2a3bf000c38&client_secret=7220f2d305asdfasdfasdf82bc3ad738b882"
20:53:11  <mbalho>:D
20:53:20  <defunctzombie>I hope that is a joke :p
20:54:02  <mbalho>also you should store passwords in plaintext because you will get more scale that way
20:54:13  <defunctzombie>yes
20:54:18  <defunctzombie>they are faster to check that way
20:58:45  <CoverSlide>i remember hacking on vbulletin, they did something pretty awesome for when people disable cookies, they keep a session in the db, and append a ?session_id=87c6fds8f7afd8s to all links.
20:58:51  * fotoveritequit (Remote host closed the connection)
20:59:10  <CoverSlide>which means link sharing also includes their session id
21:11:18  <substack>defunctzombie: inkscape + wacom tablet
21:11:33  <defunctzombie>substack: cool :) I just got a monoprice tablet
21:11:38  <defunctzombie>should arrive in a few days
21:11:46  <defunctzombie>was too cheap to get a wacom haha
21:12:28  <substack>I got my wacom for $100
21:12:35  <substack>in 2009
21:14:39  * No9quit (Ping timeout: 256 seconds)
21:16:08  * No9joined
21:27:32  * Dspriggspart
21:29:25  * chrisdickinsonquit (Excess Flood)
21:35:51  * chrisdickinsonjoined
21:38:21  * chrisdickinsonquit (Excess Flood)
21:43:18  <defunctzombie>substack: https://github.com/substack/jsonify did you ever bench this vs the builtin?
21:43:30  <substack>no
21:43:39  * AndChat648704quit (Ping timeout: 260 seconds)
21:43:46  <substack>it's just crockford's thing but as commonjs that doesn't puke all over the builtin prototypes
21:44:19  <defunctzombie>I know what it is :) but just wondering how it compares to the builtin json stuff
21:45:51  * chrisdickinsonjoined
21:48:20  <CoverSlide>builtin JSON should be faster. I don't think there's any question
22:09:45  <gozala>substack: I wrote the function analyser thing we talked about other day
22:10:02  <gozala>to spot enclosed references
22:10:03  <gozala>https://github.com/Gozala/isoscope
22:28:55  <substack>sweet!
22:39:20  * jez0990_joined
22:57:47  * lyaunzbequit (Ping timeout: 252 seconds)
23:12:57  * AndChat648704joined
23:15:28  * fotoveritejoined
23:18:57  <Raynos>ANyone have a non-jquery version of http://daniel-zahariev.github.com/jquery-textntags/ ?
23:19:27  <Raynos>mbalho: https://github.com/Raynos/push-queue
23:19:55  <Raynos>chrisdickinson: promises are good for functional style code. But gozala/reducers is >>> promises
23:21:23  * ryan_stevensjoined
23:24:42  * stlsaintquit (Ping timeout: 264 seconds)
23:25:14  <gozala>Raynos: In a way yes, although in a future they will be less so
23:25:33  <gozala>https://github.com/Gozala/reducers/pull/31
23:26:07  <gozala>chrisdickinson: Raynos: I came to conclusion that it's better to think in terms of streams then promises
23:26:18  <gozala>since promise is just a stream with single item
23:26:36  <gozala>while streams are not as easy to express in terms of promises
23:26:38  <Raynos>gozala: why?
23:27:15  * dominictarrjoined
23:27:30  <gozala>So if you treat promises as single item streams they are compose-able with rest of your stream stuff
23:28:10  <gozala>by streams I don't mean nodes stream implementation though, but a general abstraction
23:29:01  <Raynos>I mean
23:29:13  <Raynos>why will promises be less good or why are reducers less better
23:29:26  <Raynos>never mind.
23:30:34  <st_luke>defunctzombie: is localtunnel broken?
23:30:46  <defunctzombie>st_luke: shouldn't be
23:30:48  <defunctzombie>let me see
23:30:56  <st_luke>keeps crashing for me
23:31:34  <defunctzombie>I just ran it and it seemed to work
23:31:37  <defunctzombie>let me check the logs
23:34:12  * nk109joined
23:34:49  <defunctzombie>wait.. I don't have any logs hahaha
23:35:02  <defunctzombie>st_luke: that server is under heavy load right now cause of courseoff usage
23:35:10  <defunctzombie>so that might be affecting localtunnel a bit
23:36:03  <st_luke>defunctzombie: I found the issue, I was doing something weird
23:36:04  <st_luke>its all good
23:36:10  <defunctzombie>k
23:40:09  <hij1nx>st_luke: lev looking much more solid. now has a `use` command that will auto complete the same way that `require('` does so that way you dont have to type out the dirname.
23:40:43  <hij1nx>this thing has made my productivity level go way up, now i can inspect the database quickly and easily.
23:41:45  <hij1nx>also, i like autocomplete/suggest on keys since i store keys that have sha1 hashes in them
23:52:38  <Raynos>winning
23:52:49  <Raynos>hij1nx: can you make it work with levelidb through a server
23:53:13  <Raynos>just proxy commands to the client over a websocket
23:59:43  <fotoverite>grilledcast?