00:06:46  * eyeshlootchanged nick to isaacs
00:09:08  <niftylettuce>so how can responseTime be accurate if it gets calculated right after it is set? https://github.com/senchalabs/connect/blob/master/lib/middleware/responseTime.js
00:09:17  <niftylettuce>o sorry wrong channel
00:13:33  * thatguydanquit (Quit: thatguydan)
00:18:39  * st_lukejoined
00:19:18  * joshwnjjoined
00:24:17  * thatguydanjoined
00:27:18  * joshwnj`joined
00:28:28  * joshwnjquit (Ping timeout: 244 seconds)
00:38:41  * joshwnj`changed nick to joshwnj
00:46:03  <pkrumins>look guys what i just made:
00:46:06  <pkrumins>http://www.catonmat.net/download/bash-redirections-cheat-sheet.pdf
00:49:18  * thatguydanquit (Quit: thatguydan)
00:49:23  * st_lukequit (Remote host closed the connection)
00:52:33  <joshwnj>pkrumins: nice! i've really been enjoying all the bash tips on your site, thank you
00:53:10  <pkrumins>you're welcome
01:08:51  <joshwnj>pkrumins: is there a standard way to do "persistent" bash history? eg. if there are a bunch of commands I always want to appear in my history and not get cycled out?
01:13:56  * joshwnj`joined
01:16:27  * joshwnjquit (Ping timeout: 276 seconds)
01:17:49  * joshwnj`changed nick to joshwnj
01:18:18  <pkrumins>there isn't a standard way
01:18:38  <pkrumins>only workarounds
01:18:39  <mbalho>you can set the num of lines bash should story in `history`
01:18:51  <mbalho>store*
01:19:03  <mbalho>just set export HISTSIZE=9999999
01:19:09  <pkrumins>yes that's one of the workarounds
01:19:29  <pkrumins>another is to append the commands to your bash history file after each session
01:19:48  <pkrumins>.bash_logot
01:19:51  <pkrumins>.bash_logout
01:20:47  * tilgovijoined
01:23:45  * joshwnj`joined
01:25:16  * joshwnjquit (Ping timeout: 244 seconds)
03:04:14  * st_lukejoined
03:05:44  <substack>need to start augmenting npm with hella metadata for browser stuff so people stop writing these: https://github.com/twitter/bower
03:08:48  <guybrush>also component, jamjs, .., .., .., ..
03:10:11  <guybrush>i just dont understand why they dont use browserify lol.... its just not understandable
03:10:55  <guybrush>they should write another thing which does exactly the same what bfy is doing - if anything
03:12:11  <guybrush>http://jamjs.org/
03:15:52  <substack>browserify just uses npm because there are already tens of thousands of packages all up in it
03:17:01  <guybrush>well you got me with the idea, i am really into it - truly think this is the right way
03:17:12  <st_luke>ender isn't bad, it uses npm and does require the right way in the browser
03:17:15  <st_luke>but it's a build step
03:17:18  <guybrush>the implementation-detail is side issue
03:18:04  <guybrush>the problem is, big part of the community doesnt think so
03:18:48  <guybrush>i think we will end up with some thing that tj makes and everyone uses
03:19:05  <st_luke>ah yes just like express
03:19:08  <st_luke>something everybody uses
03:19:09  <guybrush>it will all be on npm still, but in some weird way
03:19:17  <substack>I'll just use the awesomest modules available
03:19:21  <substack>who cares if nobody else uses them
03:19:46  <substack>I'm running a ton of modules in production that have less than 5 stars on github
03:19:47  <st_luke>I don't particularly care if a module is popular or not
03:19:54  <substack>plenty that just have 1 star
03:20:09  <substack>but they work and have tests so that's awesome
03:20:27  <st_luke>building good software is hard enough to worry about if you're using popular components
03:20:57  <guybrush>building good stuff is hard indeed
03:21:50  <substack>there are some very reliable tricks
03:21:59  <st_luke>illusions
03:22:57  <st_luke>I had a really good time today and yesterday in the enterprise world
03:23:06  <guybrush>also, the more my stuff relies on my own stuff i.e. built by me and not reviewd by anyone else - the more i fear everything could break :PP
03:23:19  <substack>ender isn't so bad, it just have a different focus
03:23:42  <substack>I use some modules originally written for ender in browserify with no problems
03:24:09  <substack>st_luke: did you get to meet captain picard?
03:24:25  <guybrush>bwaha
03:24:26  <st_luke>substack: I wish
03:24:33  <guybrush>fighting klingons all day
03:24:40  <substack>aside: it's odd that a socialist utopia would name one of their flagship spaceships "enterprise"
03:25:21  <st_luke>substack: I think it was more tradition
03:26:01  <substack>maybe it's related to how the 24th century is obsessed with the 19th and 20th centuries
03:26:10  <st_luke>yeah
03:26:14  <st_luke>lots of boats named enterprise
03:27:52  <st_luke>I went into the office and the internet wasn't working. there was a guy surrounded by networking cables and giant switches on his laptop. he said, "I'm not really good with networking."
03:27:55  <st_luke>at least he's honest
03:27:57  <AvianFlu>an enterprise also isn't just a business
03:28:04  <AvianFlu>it's another word for a venture or endeavor
03:28:05  <AvianFlu>or a journey
03:28:10  <st_luke>AvianFlu: or a starship
03:28:14  <guybrush>in stargate the humans first ship was called prometheus, best name ever
03:29:16  <substack>LOUDBOT: whosaid
03:29:16  <LOUDBOT>substack: twentynine in #mefi on slashnet
03:30:16  <substack>blarg going to draw some pix
03:30:21  <substack>just because!
03:42:27  <mbalho>substack: requesting cats in a tree
03:42:54  <substack>http://substack.net/images/tree-traversal.png
03:43:52  <mbalho>omgz
03:45:44  <substack>turn-around time: 27 seconds
03:47:15  <mbalho>A++++++==+++ WOULD REQUEST CAT DRAWINGS AGAIN
03:49:03  <st_luke>I'm looking forward to seeing the taco conf 2013 logo
03:51:30  <mbalho>im thinking BBQONF
03:52:25  <st_luke>that would be a great regional taco conf
03:54:44  <substack>flub I need some arts
03:57:45  <st_luke>alameda-weehawken burrito tunnel?
04:06:27  <st_luke>the next city I move to must have a good burrito/taco presence
04:08:01  <AvianFlu>I recommend san diego, if that's your primary concern
04:10:01  <st_luke>also has to have the java scripts
04:10:44  <st_luke>I'm a little disappointed in the nyc java scripts right now, hopefully empire can help things a little bit
04:17:37  <substack>it's mostly about having a small core few people who are awesome and like to meet up places to hack
04:21:09  <st_luke>substack: yeah, that's what we need. There are a few meetups but they're really big and all about being `hardcore super awesome very hirable making lots of moneys` programmers
04:21:37  <mbalho>nyc java script community is lacking?
04:21:41  <mbalho>is it too bro-ey?
04:23:39  <st_luke>mbalho: no, things just seem very business oriented to me in nyc
04:23:39  <st_luke>business first
04:23:58  <substack>bros are highly correlated with businessiness
04:25:17  <mbalho>yep
04:25:43  <substack>so when I land in shanghai I've decided I'll just wander around until I find a subway station
04:25:55  <substack>might not even get off at the right stop on purpose
04:29:02  <st_luke>I did that in thailand
04:29:21  <st_luke>but it was not a subway, it was the back of a motorcycle taxi
04:29:45  <substack>sweet
04:29:51  <substack>maybe I can find a guy with a rickshaw
04:29:52  <st_luke>and I mainly got off because I realized the driver was drunk after we were driving on the sidewalk weaving through a bunch of pedestrians at a high speed
04:32:52  <st_luke>I want to go to jscamp asia, I should get job to pay for that before I leave company
04:44:57  <mbalho>substack: DID YOU USE STENOGRAPHER YET?
04:45:13  <mbalho>substack: i added a mode that works on linux: cat speech.mp3 | stenographer -t mp3
04:45:33  <mbalho>where speech is some short vocal audio file and -t specifies the input format
04:45:43  <substack>negatory
04:45:54  <substack>I will try it momentarily
04:53:47  <substack>saw this thread on the twitters http://news.ycombinator.com/item?id=4498016
04:57:06  <guybrush>not related to that bower thing, but i have to say i really hate hackernews - why do people post things there? some magically algo which sorts the posts.. wtf? i just dont understand why hackernews got so big
04:57:54  <guybrush>it like the paradise for those bro guys with only business interests
05:00:06  <st_luke>mbalho: should stenographer automatically go to `say` on os x?
05:00:34  <mbalho>it should probably parse the response text and then you could just pipe to say
05:02:01  <substack>guybrush: network effects
05:02:05  <substack>HN can draw a LOT of traffic
05:02:23  <substack>and sometimes people post about stuff that is relevant to technology
05:04:00  <st_luke>it's an ok time killer if you pretend the comments don't exist
05:06:03  <guybrush>well maybe i just see it wrong, its not a technology-blackboard - its an attention-whore-thing :p
05:09:19  <guybrush>tough i have to admit theres lots of usefull stuff on it, and sometimes you have to be an attention-whore to get important stuff out
05:09:55  <guybrush>we should post browserify-stuff regularly to make everyone use it
05:11:51  <st_luke>probably not worth it
05:12:27  <st_luke>the hn community reminds me of the people I work with that have a laugh at me wanting to build server-side components in node because it's fun and works well
05:13:39  <guybrush>working shouldnt be fun! and javascript is toy-language anyway
05:55:41  * idletomquit (Ping timeout: 268 seconds)
06:02:22  <substack>st_luke: hah yeah best to just ship code and write tiny modules that do one thing well instead
06:07:52  <mbalho>substack: is there a way to parse the samplerate from the play command? i wanna get sampl erate programatically
06:08:41  <substack>from files with framing like wav and mp3?
06:09:28  <substack>soxi
06:09:52  <substack>same as sox -i
06:10:10  <substack>oh wait seems different, pesky docs
06:10:19  <substack>anyways soxi works
06:10:42  <mbalho>sweet
06:12:40  <mbalho>does that get installed when sox gets installed/
06:13:36  <st_luke>mbalho: looks like it does through homebrew
06:13:42  <mbalho>sweet
06:39:20  <mbalho>st_luke: new version of stenographer is way more accurate
06:41:18  <mbalho>isaacs: o/
06:41:35  <mbalho>isaacs: turns out i wasnt setting the sample rate accurately so things were super messed up
06:45:22  <mbalho>isaacs: ok now i think stenographer would be useful for nodeup transcriptions
06:46:43  <st_luke>mbalho: sometimes I'm getting weird errors https://gist.github.com/3689281
06:47:39  <mbalho>yea i think it works best with between 2 - 20 words
06:49:43  <mbalho>substack: does sox have an option to slice an incoming file with beginning and end offsets in seconds?
06:49:54  <substack>mbalho: https://gist.github.com/3689295
06:50:09  <substack>yes it does
06:50:45  <mbalho>substack: it only works on mac right now cause macs have a `say` command for doing text to speech
06:50:51  <substack>I see.
06:50:59  <mbalho>substack: if you know a linux cli thing that does text to speech and can write to audio files or stdout lemme know
06:51:08  <substack>there's the flite command
06:51:11  <substack>similar to say
06:51:14  <st_luke>this is actually pretty accurate
06:51:15  <st_luke>nice
06:51:17  <mbalho>substack: you can use this syntax though: echo file.wav | stenographer -t wav
06:51:41  <st_luke>pretty sweet module
06:53:23  <substack>mbalho: there is a "splice" in sox
06:53:28  <substack>not quite sure how to use it though
06:54:09  <mbalho>substack: looks like trim will do it
06:54:14  <substack>yep just found that
06:54:18  <mbalho>substack: sox inputfile outputfile trim 0 10
06:55:01  <mbalho>substack: im gonna make it break up input audio streams into 10 second chunks and process them one at a time via the google api
07:05:19  * westonruterquit (Quit: westonruter)
07:31:38  <st_luke>http://i.imgur.com/NQ5TP.jpg
07:34:47  <mbalho>lol
07:49:36  * st_lukequit (Remote host closed the connection)
07:59:28  <substack>I think 疯狂科学 is chinese for mad science
08:00:52  <mbalho>haha
08:03:16  <substack>oh fuck yes chinese is sweet
08:03:30  <substack>you can just take each character and plug it into google translate individually
08:03:42  <substack>crazy mad branch school
08:03:53  <substack>highly modular language
08:09:47  <guybrush>my brother studies chinese, there are chinese signs all over our house - though i cant read any :p
08:10:09  <guybrush>he takes a semester abroad in october in shanghai
08:10:54  <substack>neat
08:11:03  <substack>I'll just be there on tuesday
08:11:08  <guybrush>its crazy hard to learn this language and even more to speak it
08:11:26  <substack>fake it til you make it
08:11:32  <guybrush>ha
08:12:15  <guybrush>your doing a javascript-thing in shanghai?
08:12:36  <substack>yep
08:12:41  <substack>http://hujs.org/
08:12:43  <guybrush>amazing
08:13:35  <substack>pretty much!
08:15:43  <guybrush>the website is pretty cool too
08:16:14  * joshwnj`quit (Ping timeout: 248 seconds)
08:16:42  <guybrush>wooooooah i need more money to travel to all these awesome events
08:17:36  <guybrush>do they pay all your expenses? like for all speakers
08:19:16  <guybrush>this looks like 14-16. sept is shanghai the most awesome place to be :p
08:21:10  <substack>they're paying for airfaire and hotel
08:21:40  <guybrush>sounds like a winwin-situation
08:21:48  <guybrush>they will be happy to have you for sure
08:26:39  <guybrush>wow this article is interesting! https://twitter.com/ioerror/status/245056647133818880
08:27:25  <substack>yep I read that this morning
08:28:45  <mbalho>woot! new version of stenographer supports arbitrarily long audio files
08:28:51  <mbalho>actually that is not true
08:28:58  <mbalho>right now it buffers the whole file into memory
08:29:13  <mbalho>cause writing it in a pure streaming was too hard for 1AM
08:47:51  <substack>so I now have the central theme
08:48:24  <substack>https://en.wikipedia.org/wiki/Chinese_dragon#Dragons_and_tigers
08:48:48  <substack>dragons: small unixy apps
08:49:01  <substack>tigers: giant design-patterny big apps
08:49:05  <substack>YESSSSS
08:49:51  <guybrush>"Dragon style" is used to describe styles of fighting based more on understanding movement, while "Tiger style" is based on brute strength and memorization of techniques
08:49:54  <guybrush>haha
08:50:41  <substack>best thing ever
08:50:45  <guybrush>yes i can see the correlation
08:50:53  <substack>so perfect
08:54:28  <guybrush>i really like the example they explained in the at&t video, where they search for spelling-errors by using a bunch of unix-tools
08:55:16  <guybrush>it just brings it to the point, the concept is so powerful
08:55:20  <substack>that was brian kernighan!
08:55:34  <substack>of k&r fame
09:16:00  <guybrush>is the stream-api in node going to change a lot in 0.9?
10:22:23  <substack>http://substack.net/images/mad_science_%E7%96%AF%E7%8B%82%E7%A7%91%E5%AD%A6.png
10:22:48  <guybrush>holy cow that is a nice picture ^^
13:30:01  * sorensenquit (Ping timeout: 256 seconds)
13:32:01  * chapelquit (Ping timeout: 268 seconds)
13:57:53  * chapeljoined
14:01:54  * wiwilliajoined
14:32:58  * st_lukejoined
15:21:23  <isaacs>substack: looking forward to the slides from whatever talk it is you're preparing
15:21:28  <isaacs>substack: and video
15:21:47  <pkrumins>hujs
15:22:00  <isaacs>substack: i always try to guess at what it is you're going to be getting at from the pictures that show up on twitter in the days leading up to the event
15:22:07  <pkrumins>isaacs: look what i made http://www.catonmat.net/download/bash-redirections-cheat-sheet.pdf
15:24:34  <isaacs>pkrumins: that's handy!
15:24:59  <pkrumins>sure is!
15:34:02  * westonruterjoined
16:17:12  <Altreus>pkrumins: now that's a cheat sheet I need a bunch
16:46:40  * idletomjoined
16:57:37  <st_luke>I love when people's jobs buy them conference tickets
16:58:19  <st_luke>although it's slightly reminiscent of taking an amusement park field trip during middle school but being given group work to do while you're there
17:16:07  * wiwilliaquit (Ping timeout: 260 seconds)
17:30:37  * westonruterquit (Quit: westonruter)
17:31:52  * tilgoviquit (Remote host closed the connection)
17:41:51  * wiwilliajoined
17:53:44  * tilgovijoined
17:55:41  * rannmannquit (Changing host)
17:55:41  * rannmannjoined
18:01:43  <Raynos>guybrush: the readable stream yes
18:26:24  <mbalho>substack: do you remember the conversation with creationix at summercamp about how http streams should emit 'headers' and 'data'?
18:27:37  <substack>I don't remember that part
18:28:37  <substack>+1 to an option to turn off the chunked encoding in the response stream too
18:29:18  <substack>that would make writing proxies so much easier
18:38:11  <isaacs>Raynos: just pushed readable 0.0.2 to github and npm
18:38:26  <isaacs>Raynos: still not completely happy with it
18:39:06  <isaacs>but i'll work more on it this evening.
18:39:16  <isaacs>Raynos: er, after 2 or so, that is
18:39:25  <isaacs>Raynos: if you wanna post more issues, you are once again welcome ;)
18:39:43  <isaacs>Raynos: but multipipe should work properly now, and flow() happens on nextTick
18:39:56  <isaacs>Raynos: the api surface has gotten too big, though
18:40:05  <isaacs>it should be almost nothing
18:49:48  <mbalho>isaacs: most times when i run stenographer i get "Error: write EPIPE" errors but one out of every 6 or so runs it works fine
18:50:22  <isaacs>mbalho: that means that someone's stdout closed, then you tried to write to it
18:50:29  <isaacs>mbalho: usually you can just ignore that
18:50:45  <isaacs>process.on('error', function (er) { if it's an epipe, then do nothing, otherwise throw } )
18:50:49  <mbalho>cool
18:50:51  * isaacsaway
18:57:19  * sorensenjoined
19:13:03  * dominictarrjoined
19:25:40  * tphummeljoined
19:27:50  <tphummel>hello. trying to setup a browserling tunnel. i can curl for my usage stats. i curled in my pub key. opened the tunnel. the final setup step hangs: ssh -NR 57863:localhost:8080 ...
19:27:51  <substack>dominictarr: booked my flight out of berlin, I'll be there from oct 3 to 12
19:28:07  <dominictarr>substack, sweet!
19:28:10  <substack>tphummel: it's supposed to hang!
19:28:20  <tphummel>asks for my pubkey pass. then my browserling account pass. then hangs
19:28:23  <substack>yep!
19:28:26  <substack>that means it's working
19:28:39  <tphummel>ah :-)
19:28:52  <pkrumins>it shouldnt be asking for browserling account pass though
19:28:55  <dominictarr>substack, so you hanging out in lisbon a while?
19:28:55  <substack>so now in the ui you just go to http://tunnel.browserling.com:57863/
19:29:04  <substack>dominictarr: yep
19:30:38  <tphummel>in my ui? you mean in my test code. i get "can't connect" when i hit it in chrome
19:30:47  <tphummel>apologies for nub questions
19:31:16  <pkrumins>ui = in the browser inside of browserling
19:32:07  <substack>pkrumins made a video showing how it works, could be helpful
19:32:17  <pkrumins>that video is no longer valid
19:32:23  <pkrumins>as we dont do password based auth
19:32:29  <substack>oh true
19:33:00  <substack>hmm I should use that old ssh icon in the ui, it was sweet
19:33:18  <substack>also more compact horizontally so we can fit more buttons in
19:34:55  * simcop2387quit (Excess Flood)
19:35:58  <tphummel>pkrumins, substack : thanks very much
19:36:34  * simcop2387joined
19:44:06  <dominictarr>ahh, fuck.
19:44:21  <dominictarr>I've never missed a flight. but I have booked the wrong flight.
19:44:40  <dominictarr>nearly as expensive. but at least I don't have to get up.
19:49:48  <mbalho>are you all staying at the conf hotel in lisbon? i got an airbnb
19:52:39  <dominictarr>nah, I got an airbnb
19:53:24  <mbalho>dominictarr: already booked? mine could house another probably
19:53:38  <dominictarr>yeah, already booked.
19:53:45  <mbalho>werd
19:54:25  <dominictarr>are you gonna hang out in lisbon after?
19:55:06  <mbalho>goin to helsinki
19:55:22  <substack>dominictarr: http://substack.net/images/mad_science_%E7%96%AF%E7%8B%82%E7%A7%91%E5%AD%A6.png
19:55:32  <substack>疯狂科学 -> mad science
19:55:43  <dominictarr>I saw! thats awesome
19:55:58  <dominictarr>like the bow tie.
19:56:11  <Raynos>isaacs: https://github.com/raynos/read-stream As for API service on readable I tried to create a simple API for read-stream
19:57:21  <dominictarr>Raynos, I managed to moronically book the wrong return flight for lxjs so I need to rebook anyway,
19:58:07  <dominictarr>so I can just rebook via london!
19:58:07  <pkrumins>maddo sciencoooo
19:58:11  <pkrumins>as japanese would say
19:58:16  <dominictarr>haha
19:58:22  <mbalho>haha
19:59:29  <substack>マッドサイエンス
20:00:06  <substack>http://translate.google.com/#auto/ja/%E3%83%9E%E3%83%83%E3%83%89%E3%82%B5%E3%82%A4%E3%82%A8%E3%83%B3%E3%82%B9
20:00:09  <substack>click the listen button
20:00:25  <pkrumins>haha
20:01:04  <mbalho>whoa
20:02:17  * tilgoviquit (Ping timeout: 264 seconds)
20:02:23  <rowbit>/!\ ATTENTION: (default-local) [email protected] successfully signed up for developer browserling plan ($20). Cash money! /!\
20:02:23  <rowbit>/!\ ATTENTION: (default-local) paid account successfully upgraded /!\
20:02:27  <substack>woot
20:02:52  <substack>https://ja.wikipedia.org/wiki/%E3%83%9E%E3%83%83%E3%83%89%E3%82%B5%E3%82%A4%E3%82%A8%E3%83%B3%E3%83%86%E3%82%A3%E3%82%B9%E3%83%88
20:03:09  <Raynos>dominictarr: where are you now?
20:03:21  <dominictarr>berlin.
20:03:54  <Raynos>let me know when you fly through lxjs
20:04:00  <dominictarr>I'm thinking fly to london next monday, then to lisbon on the 27th.
20:04:17  <dominictarr>that way, can go to http://lnug.org/ on the 26th
20:08:40  <substack>mbalho: phase 2 of stenography project, markov bots
20:09:04  <mbalho>substack: yep. new version of stenographer actually works pretty well
20:09:26  <mbalho>substack: so if there is a markov bot that you can pipe() text streams to that would work
20:16:00  * tilgovijoined
20:16:08  <dominictarr>Raynos, are you going to LXJS?
20:16:15  <Raynos>doubt it.
20:16:22  <Raynos>I've got no flights scheduled in the near future
20:16:36  <dominictarr>they are out of tickets anyway.
20:16:42  <Raynos>This reminds me I need to apply to talk at places.
20:16:52  <Raynos>I'd rather get a free ticket by being a speaker
20:16:56  <dominictarr>I'm thinking fly to london next monday, then to lisbon on the 27th.
20:16:57  <dominictarr>that way, can go to http://lnug.org/ on the 26th
20:17:05  <Raynos>i signed up for lnug
20:17:21  <dominictarr>I think it says they want speakers. you gotta start with meet ups.
20:17:23  <Raynos>you can probably also crash at mine for a few days
20:17:35  <Raynos>yeah I've done a few meetups
20:17:37  <Raynos>I should do more
20:18:01  <dominictarr>cool, I got other friends in london two, that I need to visit
20:18:39  <Raynos>o/
20:18:49  <Raynos>good luck scheduling your couch surfing
20:20:38  <dominictarr>thanks.
20:27:49  <dominictarr>Raynos, we had better make a visit to bletchley park!
20:27:56  <Raynos>why?
20:28:21  <Raynos>we can if you want
20:29:35  <dominictarr>you know that whole history of computer science WW2 code breaking thing.
20:30:21  <dominictarr>Raynos, read cryptonomicon
20:30:41  <Raynos>so much to read / do / etc
20:32:28  <substack>cryptonomnomnomicon
20:35:52  * tilgovi_joined
20:36:06  * tilgoviquit (Ping timeout: 240 seconds)
20:38:23  * tilgovi_changed nick to tilgovi
20:50:54  * tilgoviquit (Ping timeout: 244 seconds)
20:54:49  * wiwilliaquit (Ping timeout: 255 seconds)
20:55:39  * tphummelquit (Quit: tphummel)
20:56:38  * wiwilliajoined
20:58:22  * joshwnj`joined
21:12:10  <isaacs>mbalho: what's the event?
21:13:50  <guybrush>there was already some computer science during ww1! it had even some impact on how it went
21:13:57  <guybrush>http://en.wikipedia.org/wiki/Room_40
21:14:07  <guybrush>super awesome stuff to read
21:21:44  <substack>http://substack.net/images/unix_philosophy.png
21:21:53  <substack>UNIX哲学!
21:25:04  <pkrumins>NUXI
21:25:15  <pkrumins>http://catb.org/jargon/html/N/NUXI-problem.html
21:25:28  <substack>haha
21:29:18  * wiwilliaquit (Ping timeout: 260 seconds)
21:31:45  * tilgovijoined
21:33:12  * wiwilliajoined
21:33:43  <Raynos>dominictarr: https://gist.github.com/3692584
21:33:52  <Raynos>Am I doing it wrong if I need flatten?
21:37:42  <dominictarr>Raynos, what is the use-case?
21:37:45  * mikealjoined
21:37:51  <dominictarr>sometimes there is a need for flatten.
21:38:20  <dominictarr>seems reasonable, but if you could do without it then I guess you should.
21:41:14  <Raynos>dominictarr: https://gist.github.com/444b921e23fa9e4f5683
21:41:24  <Raynos>mapping a stream of event emitters into a stream of streams and flattening that into a single stream
21:45:23  <dominictarr>I draw the line at passing live objects through streams. assert.deepEqual(data, JSON.parse(JSON.stringify(data))) should always pass.
21:45:39  * tilgoviquit (Ping timeout: 260 seconds)
21:46:16  <substack>serializable objects seems legit
21:46:18  <dominictarr>where do the event-emitters come from?
21:47:10  <isaacs>dominictarr: i'm thinking about your "suck stream" approach.
21:47:16  <dominictarr>serializable objects could be converted to text streams my a script of low complexity, and therefore they are the same thing.
21:47:38  <isaacs>dominictarr: it's somewhat similar to irakli's "reducible" thing
21:47:47  <dominictarr>link?
21:47:58  <isaacs>which i kinda dig, but is too big a diff from what we have, and from tcp and from parsers.
21:48:22  <isaacs>https://gist.github.com/2656978
21:48:59  <isaacs>dominictarr: so, ultimately, the goal is to make streams be a very clear mapping to the semantics of sockets
21:49:10  <isaacs>i think history has spoken, and this is the correct way to do IO
21:49:28  <isaacs>fs.ReadStream and fs.WriteStream is a bandaid over the biggest mistake in unix
21:50:44  <Raynos>dominictarr: passing live objects in it is fun. Because you can pass DOM elements down it ( https://gist.github.com/b631ff61d75597c2b6b1 )
21:52:15  <dominictarr>Raynos, that is a reasonable border line case. if the have references (listeners) then it could be dangerous.
21:52:46  <dominictarr>isaacs, what is "the biggest mistake in unix" ?
21:54:26  <Raynos>why are listeners dangerous?
21:55:20  <dominictarr>Raynos, well, it just explodes the complexity, potentially.
21:55:45  <Raynos>dominictarr: It does, I'm trying to think about how to implement MovableList & ElementList using only streams
21:55:49  <Raynos>this reminds me I should ask gozala
21:55:52  <dominictarr>I've seen some streaming type constructs that go too far
21:56:13  <dominictarr>like, trying to do recursion and loops in streams
21:57:25  <Raynos>loops in streams is easy
21:57:38  <Raynos>https://github.com/Raynos/lazy-reduce-stream#example
21:57:48  <Raynos>Just apply a reduction over a stream of data and use the accumulator as state
21:58:41  <dominictarr>Raynos, look at this for an example http://architects.dzone.com/articles/nature-pipes
21:59:12  <dominictarr>see "loop" pipes. when you add stuff like that you havn't made anything simpler.
21:59:39  <dominictarr>the point of streams is that it's NOT a universal computation.
22:01:21  <dominictarr>isaacs, one thing that concerns me with readable-stream is that it may lead to having StreamReader, StreamWriter classes like in java.
22:04:09  * wiwilliaquit (Ping timeout: 252 seconds)
22:05:15  <dominictarr>Raynos, although, I understand haskell has done stuff like this http://en.wikipedia.org/wiki/Reactive_programming
22:05:24  <dominictarr>it's quite a lot like streams.
22:05:57  <Raynos>dominictarr: I'm modeling reactive programming using streams as an implementation detail
22:06:14  <substack>time-invariant streams would be a fun experiment
22:06:52  <Raynos>time-invariant?
22:07:11  <substack>time as a parameter instead of listening at a particular time
22:07:24  <substack>like couchdb changes with since
22:07:41  <substack>that's what a lot of frp is about
22:08:39  <substack>for 3d simulations at least which is the domain I'm most familiar with for using reactive programming techniques
22:11:10  <dominictarr>substack, scuttlebutt is basically like that.
22:11:18  <substack>neat
22:11:40  <dominictarr>but you pass in a {} of node: timestamps
22:11:54  <dominictarr>(well, I mean, that is all handled under the hood)
22:13:57  <Raynos>I'm actually starting to think that named streams are better then event emitters
22:14:22  <Raynos>emitter.on("foo", handler) vs thing.foo.pipe(to(handler))
22:14:56  <Raynos>Raynos: Have you tried implementing event emitters using streams and then implement streams using event emitters. Turtles all the way
22:26:10  * tphummeljoined
22:30:42  <dominictarr>Raynos, maybe you need a simpler stream interface that doesn't depend on event emitters.
22:30:54  <dominictarr>maybe it could just use functions more directly.
22:31:07  <Raynos>I was having a joke :P
22:31:27  <dominictarr>haha, okay.
22:31:47  <Raynos>although I can't think of how to do it without dart style event things
22:31:48  <dominictarr>wasn't sure. seemed like the sort of thing you might do.
22:32:21  <dominictarr>maybe you could just have a listener that was just a function (data) { ... }
22:32:30  <dominictarr>and when data === null, that means end.
22:33:19  <dominictarr>function (data, emit) { emit(data) }
22:33:26  <dominictarr>would be a through stream.
22:33:42  <dominictarr>then piping would just be a curry.
22:33:48  <Raynos>I'm going to stick to node streams for now
22:34:17  <dominictarr>probably a good idea.
22:34:31  <Raynos>until I am convinced that my abuse of them is stupid
22:39:24  <substack>小软件 vs 大软件
22:49:01  * tilgovijoined
22:56:21  <Raynos>isaacs: every writable stream has to implement destroy because of 0.8 Stream#pipe :(
22:56:35  <Raynos>This dirties my clean 0.9 writable streams that have just write / end
22:56:41  <isaacs>Raynos: we're gonna remove that in 0.9
22:57:00  <isaacs>Raynos: ie, the default stream#pipe already only calls destroy() if it's there, and not if it isn't
22:57:08  <Raynos>isaacs: really?
22:57:42  <isaacs>Raynos: no, not really.
22:57:45  * isaacsjust checked
22:57:52  <isaacs>Raynos: bit it should.
22:58:02  <Raynos>Can we get that into 0.8.x ?
22:58:22  <Raynos>I'll even make the PR with tests and docs!
22:59:10  * joshwnj`quit (Remote host closed the connection)
23:39:03  * _sorensenjoined