| 00:03:54
| <Raynos> | Streams are a framework |
| 00:06:52
| * dominictarr | joined |
| 00:07:21
| <maxogden> | theres a difference between a single prototype and a framework |
| 00:09:34
| <dominictarr> | hmm, testing it nod. |
| 00:10:26
| <dominictarr> | every application contains an implicit framework. like every building contains a frame. |
| 00:10:48
| <dominictarr> | a framework is just when you refactor the structure out of an application. |
| 00:17:37
| <dominictarr> | Raynos, my mistake, it wasn't a bug. anyway there is a test for it now... |
| 00:18:43
| <Raynos> | dominictarr: ? what bug ? |
| 00:19:10
| <Raynos> | maxogden: I mean framework as in a framework gives you like a structure thing |
| 00:19:15
| <dominictarr> | Raynos, I thought it didn't emit 'end' according to allowHalfOpen. |
| 00:19:17
| <Raynos> | streams dominate my structure |
| 00:19:52
| <Raynos> | maxogden: better yet. A framework gives you a cohesive way to put all your pieces together and they work well with each other. Streams give you a cohesive way to put all your pieces together and they work well with each other |
| 00:20:43
| <dominictarr> | streams are like the knobs on duplo blocks. |
| 00:21:11
| <Raynos> | :D |
| 00:29:01
| <dominictarr> | templates, html, all that stuff is like coloring in with crayons. |
| 00:32:53
| <dominictarr> | Raynos, maxogden I wrote this today: https://github.com/dominictarr/stream-spec/blob/master/api-style.markdown |
| 00:33:09
| <dominictarr> | ^ my thoughts on the best patterns for streamy apis. |
| 00:34:06
| <tanepiper> | anyone know a good xml -> JSON parser? |
| 00:37:12
| <Raynos> | dominictarr: https://github.com/Raynos/auth-stream#example-client |
| 00:37:19
| <Raynos> | You say passing a stream to a function is bad |
| 00:37:27
| <Raynos> | how would you recommend refactoring that client |
| 00:38:09
| <dominictarr> | what does Auth to the stream? |
| 00:38:24
| <dominictarr> | does it write/read data to/from it? |
| 00:38:51
| <Raynos> | it writes a login message |
| 00:38:55
| <Raynos> | and reads the error / success message |
| 00:39:00
| <Raynos> | then it turns into a dumb pass through |
| 00:40:07
| <Raynos> | you can always manually write the login message |
| 00:40:13
| <Raynos> | I should add an example of doing just that |
| 00:40:35
| <Raynos> | https://github.com/Raynos/auth-stream#example-server the server also decorates a stream |
| 00:40:38
| <dominictarr> | it looks like there is a bug there. |
| 00:41:06
| <Raynos> | i dont think there is, but probably there is a bug somewhere in my code right now |
| 00:41:32
| <dominictarr> | you are depending that your chunk with the auth message remains intact. |
| 00:41:59
| <Raynos> | dominictarr: yes I am. that is a bug. I don't know how to handle that other then dont write to it until its open |
| 00:42:07
| <dominictarr> | like, if tcp decides to chuck it larger, then it would break. |
| 00:42:36
| <dominictarr> | basically you are just making a custom header. |
| 00:43:02
| <dominictarr> | you just need a way to delimit the end of the header. |
| 00:43:18
| <dominictarr> | like, match the first line. |
| 00:43:35
| <Raynos> | i see |
| 00:44:03
| <dominictarr> | here is some code that does that https://github.com/dominictarr/badass/blob/master/index.js#L26-52 |
| 00:46:01
| <substack> | I use jsonstream all the time |
| 00:46:48
| <substack> | selection synatx is so nice |
| 00:48:23
| <dominictarr> | different strokes for different folks... YAY FOR MODULES |
| 00:48:50
| <Raynos> | i hate debugging streams |
| 00:48:56
| <Raynos> | I have no idea wtf is going on |
| 00:52:55
| <dominictarr> | I like to try and make them so that they work synchronously, if you use them like that. |
| 00:53:19
| <dominictarr> | it makes it much easier to be exact about orders when testing. |
| 00:53:31
| <substack> | dominictarr: also reading your api doc, emit-stream could use pipes conceivably |
| 00:53:36
| <dominictarr> | it's mainly about getting the orders about events right. |
| 00:54:07
| <substack> | var ev = stream.pipe(emitStream()) |
| 00:54:10
| <dominictarr> | you could go emitstream().pipe(stream) |
| 00:54:13
| <substack> | ev.on('ping', fn) etc |
| 00:54:37
| <substack> | yes |
| 00:55:01
| <substack> | I do like the symmetry of a function that toggles the type of the argument though |
| 00:55:06
| <dominictarr> | but you'd still need to pass in the eventEmitter though |
| 00:55:19
| <substack> | yep in order to hack at .emit |
| 00:55:26
| <dominictarr> | yeah, I think in this case, the symmetry is a better win. |
| 00:55:41
| <dominictarr> | because it's consistent with it's own idea of itself. |
| 00:55:52
| <dominictarr> | and _that_ trumps everything. |
| 00:56:16
| <substack> | it's easiest to explain the api this way |
| 00:56:29
| <substack> | if you give it an emitter, it gives you a stream and vice-versa |
| 00:56:41
| <substack> | all you need to know to start using it |
| 00:56:43
| <dominictarr> | being consistent with your own idea of your self is like, the key to good designs in anything. |
| 00:57:39
| <substack> | versus if you give it an emitter it returns a readable stream but otherwise it returns a writable stream that is also an event emitter |
| 00:57:53
| <substack> | the second one is harder to keep in my head for some reason |
| 00:58:05
| <dominictarr> | agreed. |
| 00:58:31
| <dominictarr> | it's deff more elegant this way. |
| 00:58:50
| <substack> | aha I think I know why |
| 00:59:14
| <substack> | usually explaining things in terms of streams is more terse because you get to talk at a higher level |
| 00:59:37
| <substack> | but in this case because the concept challenges some fundamental precepts of streams it's not as good a fit |
| 00:59:46
| <Raynos> | auth-stream will need some work |
| 01:00:05
| <Raynos> | to be able to pipe them |
| 01:00:50
| <substack> | in the case of duplex-stream there's just no other way to do that |
| 01:02:18
| <substack> | it's pretty crazy to think about how little there is in the actual streaming code in core but how wide the applications are for consistent interface design |
| 01:03:03
| <dominictarr> | that is a very good omen. |
| 01:03:18
| <substack> | a prototype function with a single method pipe that evolved from sys.pump() |
| 01:03:44
| <dominictarr> | it's a pretty old idea though |
| 01:04:18
| <substack> | which was itself just a way of handling backpressure logic with .write() and 'drain' on emitters that did the stream-esque 'data'/'end' thing |
| 01:04:35
| <substack> | yep it's nearly identical to how the shell does pipes |
| 01:05:48
| <substack> | what are those jackhammers even doing in the middle of the street |
| 01:07:53
| <dominictarr> | somewhere I found a memo where some guy in the 1958 is like 'we need some sort of data processing thing that we and snap programs together like a garden hose" |
| 01:10:37
| <dominictarr> | here it is http://cm.bell-labs.com/who/dmr/mdmpipe.html |
| 01:11:03
| <dominictarr> | Douglas McIlroy |
| 01:14:52
| <dominictarr> | I'm off to bed! catch you guys later! |
| 01:15:09
| * dominictarr | changed nick to dominic_zzzzzzzz |
| 01:18:01
| <substack> | "Doug has been explicit in saying that he very nearly exercised managerial control to get pipes installed." |
| 01:18:41
| <substack> | using this in my lisbon talk! |
| 01:19:48
| * dominic_zzzzzzzz | quit (Ping timeout: 276 seconds) |
| 01:24:13
| <substack> | jesusabdullah: the 3rd mouse's name is Higgs |
| 01:24:25
| <substack> | on account of being elusive |
| 01:25:50
| <substack> | the the trap is the the large mousedrop collider |
| 01:25:55
| <substack> | s/the the/the/ |
| 01:33:33
| * substack | bikes up to berkeley to hack up a code shape module |
| 01:39:29
| * ryan_stevens | quit (Quit: Leaving.) |
| 02:02:38
| * substack | still at home |
| 02:20:28
| * ryan_stevens | joined |
| 02:34:17
| * mikeal | quit (Quit: Leaving.) |
| 03:48:11
| <substack> | isaacs: neat how the new npm site shows downloads |
| 03:51:46
| <Raynos> | substack: graph stream is super sweet |
| 03:53:27
| <substack> | a module to do line graphs would be rad too |
| 04:08:28
| <Raynos> | yeah |
| 04:08:31
| <Raynos> | I first need |
| 04:08:34
| <Raynos> | to write a cached event emitter |
| 04:08:39
| <Raynos> | and open source this login widget |
| 04:08:56
| <maxogden> | see also https://github.com/tmcw/stream-statistics |
| 04:46:57
| * blakmatrix | part |
| 05:37:11
| * ryan_stevens | quit (Quit: Leaving.) |
| 06:16:57
| * AvianFlu | joined |
| 06:34:55
| * timoxley | joined |
| 06:49:44
| * timoxley | quit (Quit: Computer has gone to sleep.) |
| 06:50:14
| * timoxley | joined |
| 08:15:27
| <substack> | maxogden: sweet |
| 08:15:39
| <substack> | I like streams. |
| 08:43:38
| <Raynos> | Im going to build a stream game this weekend |
| 08:43:42
| <Raynos> | a bullet is just a stream |
| 08:43:52
| <Raynos> | when you want it to hit someone you pipe it into a human |
| 08:44:27
| <Raynos> | that doesnt work nor does it make any sense |
| 08:45:30
| * timoxley | quit (Quit: Computer has gone to sleep.) |
| 09:10:41
| * AvianFlu | quit (Quit: AvianFlu) |
| 09:14:37
| * mikeal | joined |
| 09:17:55
| * timoxley | joined |
| 09:34:36
| <substack> | I wish esprima was streaming |
| 09:34:43
| <substack> | no time for that right now though |
| 10:01:16
| * dominic_zzzzzzzz | joined |
| 10:12:18
| * dominic_zzzzzzzz | quit (Ping timeout: 268 seconds) |
| 10:19:33
| <substack> | ls |
| 10:19:35
| <substack> | haha |
| 10:40:46
| * dominic_zzzzzzzz | joined |
| 10:59:30
| * dominic_zzzzzzzz | changed nick to dominictarr |
| 11:39:50
| <substack> | https://github.com/substack/code-art |
| 11:40:08
| <substack> | for my jsconf.eu talk |
| 11:48:15
| <substack> | ok that part done, now hacking on ci |
| 11:59:47
| * AvianFlu | joined |
| 15:17:19
| * wiwillia | joined |
| 15:27:21
| * st_luke | joined |
| 16:00:25
| * st_luke | quit (Remote host closed the connection) |
| 16:09:32
| * AvianFlu | quit (Quit: AvianFlu) |
| 16:48:51
| * dominictarr | quit (Ping timeout: 244 seconds) |
| 17:42:30
| * AvianFlu | joined |
| 17:50:21
| * AvianFlu_ | joined |
| 17:52:24
| * AvianFlu | quit (Ping timeout: 244 seconds) |
| 17:52:24
| * AvianFlu_ | changed nick to AvianFlu |
| 18:17:35
| * mikeal | quit (Quit: Leaving.) |
| 18:35:26
| * mikeal | joined |
| 18:54:54
| * AvianFlu | quit (Ping timeout: 265 seconds) |
| 19:01:54
| * wiwillia | quit (Ping timeout: 276 seconds) |
| 19:12:25
| * dominictarr | joined |
| 19:26:40
| * wiwillia | joined |
| 19:36:46
| * mikeal | quit (Quit: Leaving.) |
| 20:05:14
| * wiwillia | quit (Ping timeout: 264 seconds) |
| 20:09:01
| <dominictarr> | wow! hey guys check out this place I found on airbnb |
| 20:09:07
| <dominictarr> | https://www.airbnb.com/rooms/542031 |
| 20:21:21
| * ryan_stevens | joined |
| 20:25:40
| <jesusabdullah> | substack: caught mouse 4/? (already released into the wild) |
| 20:31:00
| * AvianFlu | joined |
| 20:40:31
| * timoxley | quit (Quit: Computer has gone to sleep.) |
| 21:03:35
| * wiwillia | joined |
| 21:28:08
| * dominictarr | quit (Ping timeout: 276 seconds) |
| 22:10:10
| <rook2pawn> | that looks like a very nice box |
| 22:14:08
| <Raynos> | dominictarr: I'll make a PR later to turn crdt.Row into a delta-stream ( http://github.com/Raynos/delta-stream ). I'll give a demo of why that is super awesome later as well |
| 22:51:22
| * AvianFlu_ | joined |
| 22:54:36
| * AvianFlu | quit (Ping timeout: 276 seconds) |
| 22:54:37
| * AvianFlu_ | changed nick to AvianFlu |
| 22:58:28
| * AvianFlu_ | joined |
| 23:01:53
| * AvianFlu | quit (Ping timeout: 272 seconds) |
| 23:01:53
| * AvianFlu_ | changed nick to AvianFlu |
| 23:18:19
| <Raynos> | Now streaming interface: https://github.com/Raynos/data-bind#example |
| 23:34:00
| * mikeal | joined |
| 23:51:33
| * st_luke | joined |