00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:07  * ircretaryjoined
00:03:11  * intabulasjoined
00:17:43  * defunctzombiechanged nick to defunctzombie_zz
00:18:25  * defunctzombie_zzchanged nick to defunctzombie
00:19:42  * defunctzombiechanged nick to defunctzombie_zz
00:34:35  * defunctzombie_zzchanged nick to defunctzombie
00:35:00  * intabulasquit (Remote host closed the connection)
00:36:25  * gwenbellpart
00:36:57  <mbalho>substack: simplex noise with deterministic seeds http://maxogden.github.com/voxel-simplex-terrain/?seed=0.4658187120221555
00:38:39  <mbalho>substack: i think my math for world width might be off cause its not generating voxels in the entire world AND the api needs a little work (convert to use prototypes) but its cool nonetheless
00:39:41  <substack>mbalho: I see you're inlining explode() https://github.com/maxogden/voxel-simplex-terrain/blob/master/demo.js#L54
00:39:53  <substack>you can just var explode = require('voxel-debris')(game) now
00:40:17  <substack>then you don't need createDebris() either
00:41:25  <substack>oh and voxel-debris does the setBlock() call so you only need to if (erase) explode(pos)
00:47:30  * st_lukejoined
00:57:07  * nk109quit (Quit: Computer has gone to sleep.)
00:59:38  * defunctzombiechanged nick to defunctzombie_zz
01:01:29  * defunctzombie_zzchanged nick to defunctzombie
01:09:44  * defunctzombiechanged nick to defunctzombie_zz
01:19:14  * mikealquit (Quit: Leaving.)
01:29:09  * defunctzombie_zzchanged nick to defunctzombie
01:32:26  <jez0990_>mbalho: I was just playing around with the collision detection... trying to figure out if I can make voxelAtPosition return the voxel coordinates instead of just a boolean that way you potentially (or at least I'm hoping) don't have to mess around with velocities and can just set the yawObject to just on/outside the voxel's boundaries and the velocities to 0
01:32:35  <jez0990_>if that makes sense?!
01:33:25  <jez0990_>but anyway, I just looked at chunker.js and it looks like a bit of effort to work it out, so I'm giving up for tonight :P
01:38:44  * gwenbell_quit (Read error: Connection reset by peer)
01:42:28  * AndChat648704joined
01:49:26  * mikealjoined
01:58:53  * intabulasjoined
01:59:57  * mikealquit (Ping timeout: 255 seconds)
02:02:26  * intabulasquit (Remote host closed the connection)
02:07:03  * intabulasjoined
02:07:04  * intabulasquit (Remote host closed the connection)
02:07:50  * defunctzombiechanged nick to defunctzombie_zz
02:20:02  * dominictarrjoined
02:20:06  <dominictarr>substack: mbalho chrisdickinson it possible to use a animated GIF as a texture in webGl?
02:25:39  * mikealjoined
02:25:59  <substack>dominictarr: testing
02:26:41  <dominictarr>I'm thinking 3d html
02:26:49  <dominictarr>3d 90's html
02:27:02  <dominictarr>where each page is a room
02:27:07  <substack>or an in-game web browser
02:27:12  <dominictarr>and then links are portals to other rooms
02:27:15  <substack>hah
02:27:27  <dominictarr>HOW THE WEB SHOULD HAVE BEEN!
02:27:27  <LOUDBOT>SCREEEAAAAM FOR YOUR CREEEEAAM
02:32:47  <substack>dominictarr: it appears that animated gifs don't work out of the box
02:32:52  <AndChat648704>dominictarr: (this is jez0990 under a cryptic pseudonym) three.js css3d renderer composited with webgl in the background is the answer. I'm building something like that right now :)
02:33:27  <dominictarr>css3d?
02:33:49  <AndChat648704>check out mrdoob's periodic table demo
02:35:03  <dominictarr>will do
02:44:10  * tilgovijoined
03:00:49  * st_lukequit (Remote host closed the connection)
03:16:36  * tilgoviquit (Remote host closed the connection)
03:21:44  <mbalho>dominictarr: there is a css renderer that does animated gifs but gpus dont support them natively and the css renderer wouldnt work with our voxel enging right now
03:24:53  <mbalho>jez0990_: check out Chunker.prototype.voxelVector, it converts screen coords to voxel coords
03:25:31  * dominictarr_joined
03:26:19  <mbalho>jez0990_: (relative to the current chunk)
03:26:26  * dominictarrquit (Ping timeout: 255 seconds)
03:26:27  * dominictarr_changed nick to dominictarr
03:32:41  * tilgovijoined
03:47:53  <mbalho>substack: smooth world http://maxogden.github.com/voxel-simplex-terrain/?seed=01388834234
03:48:05  <mbalho>substack: (chunk distance of 3 so the world takes a little longer to build)
03:48:35  <mbalho>substack: same world with rocks http://maxogden.github.com/voxel-simplex-terrain/?seed=21388834234
03:48:57  <mbalho>substack: same world with pillars http://maxogden.github.com/voxel-simplex-terrain/?seed=61388834234
03:49:18  <mbalho>(reverse engineering this demo http://29a.ch/sandbox/2012/voxelworld)
04:08:57  * dominictarrquit (Quit: dominictarr)
04:31:06  * tphummelquit (Quit: tphummel)
04:42:13  <chrisdickinson>mbalho: you'd probably have to resend the gif each frame.
04:42:13  <chrisdickinson>but you can definitely use a video element as a texture source, just like you'd use an image
04:42:16  <chrisdickinson>(again, you have to resend it each frame)
04:58:46  <mbalho>chrisdickinson: oooh ye agood point
05:02:37  <Raynos>blargh
05:02:59  * dominictarrjoined
05:03:04  * ITproquit
05:07:00  * shuaibquit (Ping timeout: 248 seconds)
05:15:26  <mbalho>substack: paste this into the console of a running voxel game https://gist.github.com/4499643
05:40:15  <mbalho>chrisdickinson: in node if you do require('interact') it throws ReferenceError: document is not defined
05:42:34  <mbalho>chrisdickinson: in interact/node_modules/drag-stream/node_modules/domnode-dom/lib/readable.js
05:43:55  <mbalho>chrisdickinson: i am not trying to use it in node i just want to be able to require it without it breaking
06:05:45  <Raynos>Weekend project idea!
06:06:03  <Raynos>Build a custom distributed db on top of leveldb & map-reduce
06:06:14  <Raynos>i wants to build one
06:07:10  <chrisdickinson>mbalho: that's domnode-dom
06:08:47  <chrisdickinson>fixing
06:10:40  <chrisdickinson>mbalho: published a fix to domnode-dom; fresh installs of interact will pick up that fix
06:14:16  <chrisdickinson>also, question: would you prefer a polling api for "actions" (via the keyboard) or a streaming api?
06:14:39  <chrisdickinson>or does this exist?
06:15:31  <Raynos>chrisdickinson: what do you mean?
06:15:48  <chrisdickinson>so, ideally there's a disconnect between "keyup / keydown" and "game action"
06:15:53  <chrisdickinson>so you can rebind keys to different actions
06:16:47  <chrisdickinson>so there should be an api where you can set key -> action bindings, and when those keys go up or down you get events if that key is bound to an action
06:16:48  <Raynos>ok so whats an action?
06:17:01  <chrisdickinson>"walk forward", "turn right", "sidestep right"
06:17:04  <chrisdickinson>for example.
06:17:51  <Raynos>well theres two types of action inputs based on keyboard
06:17:56  <Raynos>boolean and event
06:18:12  <chrisdickinson>hm?
06:18:17  <Raynos>like crawling for example
06:18:25  <Raynos>can be set to true if ctrl is held
06:18:32  <Raynos>and then set to false if its released
06:18:44  <Raynos>or it can be toggled between true / false when capslock is pressed
06:18:47  <chrisdickinson>ah
06:18:49  <Raynos>where as move forward
06:18:58  <Raynos>is a series of move forward events
06:19:12  <chrisdickinson>well, no, move forward would be more of a "wantstomoveforward = true"
06:19:16  <chrisdickinson>but i see what you're saying
06:19:26  <Raynos>i dont think like that
06:19:27  <chrisdickinson>like, "switch weapon" for example would be a "tap"
06:19:32  <Raynos>I think moving forward is a series of events
06:19:36  <Raynos>jumping is also a series of events
06:19:44  <Raynos>even though jumps are triggered on keydown instead of keypress
06:19:59  <chrisdickinson>the controller shouldn't be the mediator of the object physics, though.
06:20:13  <Raynos>doesnt matter
06:20:21  <Raynos>im not saying it is
06:20:34  <Raynos>if you want to translate move forward into whilst held apply delta x at 60 fps
06:20:52  <Raynos>then just sample the events at 60 fps and map to delta
06:21:01  <chrisdickinson>but you don't know your fps.
06:21:04  <chrisdickinson>necessarily.
06:21:08  <Raynos>doesnt matter
06:21:10  <mbalho>chrisdickinson: one one hand i like an api like this https://github.com/madrobby/keymaster
06:21:15  <Raynos>im saying that its out of scope
06:21:24  <Raynos>all your action api should provide
06:21:25  <mbalho>key('ctrl+r', function()
06:21:31  <Raynos>is a representation of boolean state
06:21:38  <Raynos>and a representation of streaming events
06:21:49  <mbalho>chrisdickinson: but i also want keyboard.on('w', moveForward)
06:21:49  <Raynos>thats how i would think of it
06:22:11  <mbalho>chrisdickinson: or keyboard.createReadStream('w').pipe(controls.forward)
06:22:29  <mbalho>chrisdickinson: jsut an event emitter is fine i think
06:22:39  <mbalho>chrisdickinson: since keyboard events dont really end
06:22:46  <mbalho>chrisdickinson: or begin. they are momentary
06:22:53  <mbalho>#metaphysicalstatements
06:23:12  <chrisdickinson>haha
06:23:13  <Raynos>var move = action({ type: "event", key: "w" })
06:23:26  <chrisdickinson>well, they do and don't -- "held" vs. "tapped" -- switch weapon could be a tap
06:23:48  <Raynos>var jumps = action({ key: "space", type: "tapped" })
06:23:53  <chrisdickinson>you don't want to bind actions tightly to keys
06:24:04  <chrisdickinson>because of things like, well, what if someone wants to rebind keys? or what if someone has a gamepad?
06:24:07  <Raynos>var crouched = action({ key: "ctrl", type: "held" })
06:24:21  <Raynos>chrisdickinson: then allow it to change
06:24:29  <Raynos>key: key
06:24:33  <Raynos>or key: [...]
06:25:03  <Raynos>map(sample(move, 60), function () { return { x: 1 } })
06:25:42  <Raynos>jumps are special because you want to disallow jumps from occurring until the last jump finished
06:25:44  <mbalho>ooh for the action() thing in raynos' example that would be a good plac to use a stream
06:25:55  <Raynos>mbalho: you mean gozala/reducers
06:26:38  <Raynos>fold(jumps, function () { return jump(...); /* apply backpressure */ })
06:26:43  <mbalho>keyboard.on({ key: "ctrl", type: "}, function(stream) { game.emit('crouch', stream) })
06:27:02  <Raynos>you do actually want a stream for jumps
06:27:13  <Raynos>so that the jump logic can emit("drain") when the jump finishes
06:27:29  <Raynos>or you use some other construct with well defined back pressure
06:28:08  <mbalho>chrisdickinson: so i guess for actions with beginnings and endings (like holding down a modifier key) a stream would be useful
06:28:17  <mbalho>chrisdickinson: otherwise just an event emitter is good
06:28:37  <mbalho>chrisdickinson: though streams around modifier keys might be overengineering
06:28:50  <Raynos>well you want backpressure!
06:29:05  <Raynos>or is backpressure not a good way to model disallowing double jumping
06:29:11  <mbalho>function ctrlToggle (ev) { erase = !ev.ctrlKey }; window.addEventListener('keyup', ctrlToggle); window.addEventListener('keydown', ctrlToggle)
06:29:15  * chrisdickinsonwrites a lil' api..
06:29:18  <mbalho>that is what we use right now in voxel games
06:29:38  <mbalho>then you put an if statement in your click handler
06:29:41  <mbalho>that checks if it should erase or not
06:29:43  <chrisdickinson>Raynos: backpressure might be good there.
06:29:44  <mbalho>nice and simple
06:29:45  * mikealquit (Quit: Leaving.)
06:29:50  <mbalho>(cant get much simpler)
06:30:11  <Raynos>but its stateful!
06:30:17  <Raynos>The lisp demons will consume you
06:30:30  <mbalho>everything is a nail
06:31:12  * mikealjoined
06:34:45  <mbalho>chrisdickinson: woot now voxel-server runs voxel-engine in node and generates the voxel terrain from a seed and then when clients connect they get sent game config including the seed and generate the exact same world
06:34:55  <chrisdickinson>woo
06:35:22  <mbalho>now im gonna use scuttlebutt to sync edits
06:43:36  <chrisdickinson>hmm, hm, hm
06:43:36  <chrisdickinson>https://gist.github.com/0c57ff97169fdf630980
06:44:28  <chrisdickinson>yeah, that's not good at all
06:44:38  <chrisdickinson>there are some bits i like, but it's a little.. heavy
06:45:19  <chrisdickinson>i definitely want actions to be streaming, so that you can pipe 'em to a replay file or database
06:46:35  <chrisdickinson>but you really can't deal with the actions as they come in as events, since you need to mediate them with the time delta between frames
06:46:53  <chrisdickinson>(or else players with faster computers go faster, players with slower computers go slower, or problems like that.)
06:47:35  <chrisdickinson>though for some events (like switch weapon) you don't care about dt -- it's completely instantaneous
06:50:54  * owen1quit (Ping timeout: 255 seconds)
06:57:49  <Raynos>that api is strange
06:59:35  * st_lukejoined
07:04:12  * owen1joined
07:16:45  * tilgoviquit (Read error: Connection reset by peer)
07:27:37  <gozala>Raynos: mbalho backpressure for user event streams is stupid
07:27:56  <gozala>user creates events weather you're ready to handle it or not
07:28:22  <gozala>so trying to back pressure mouse events or keyboard events make no sense
07:28:54  <gozala>you should either skip if not ready to handle or queue and handle once ready
07:28:56  <dominictarr>gozala: need a thing to give the user a electric shock, so they slow down
07:29:16  <gozala>or yeah what dominictarr suggested :)
07:29:18  <dominictarr>or you could just respond slower visually… that would probably slow them down
07:29:26  <gozala>write proposal to w3c :)
07:29:49  <gozala>so browsers implement, meanwhile we can polyfill
07:30:01  <dominictarr>and lobby for a law that all computers have a kibble dispensor
07:30:06  <gozala>dominictarr: that's queueing
07:30:34  <dominictarr>then you have both positive and negative feedback
07:32:27  <Raynos>gozala: it's not stupid
07:32:39  <Raynos>gozala: it's just a convenient way to say "drop this data until im ready"
07:32:58  <Raynos>I meant I want a way to skip
07:33:01  <Raynos>how would you skip?
07:33:02  <gozala>Then just say that instead
07:33:31  <gozala>takeWhen(events, behaviour)
07:33:44  <gozala>behaviour is stream representing state
07:33:51  <gozala>weather events are taken or not
07:33:53  <Raynos>how do you do that in non stateful fashion :p
07:34:14  <Raynos>i guess it doesnt matter
07:35:07  <Raynos>the point is consumer of spaces (jump logic) needs to tell stream someone to drop space messages
07:36:07  <gozala>You basically end up with state machine
07:36:33  <gozala>and have jumpable state
07:36:41  <gozala>when in that state then you get jump events
07:36:55  <gozala>I wrote these functions
07:37:07  <gozala>forgot the name of lib :(
07:37:48  <gozala>Raynos: here https://github.com/Gozala/transducer
07:37:58  <gozala>Also take a look at http://elm-lang.org/
07:38:05  <gozala>that's exactly what it does
07:38:20  <gozala>http://elm-lang.org/blog/games-in-elm/part-0/Making-Pong.html
07:39:49  <gozala>That post describes very well streaming state machine
07:40:02  <gozala>and how different inputs combined
07:41:04  * dominictarrquit (Quit: dominictarr)
07:41:47  <gozala>Trying to represent all of that via backpressure will just make things less obvious
07:42:07  <gozala>IMO anyway
07:44:03  <gozala>Raynos: so for example you will have signal jumping
07:44:26  <gozala>which yields true on jumps and false once jump is finished
07:44:50  <gozala>then you get jump events you need to react as
07:45:58  <gozala>jumpEvents = takeWhen(jumpKey, complement(jumping))
07:49:24  <Raynos>complement?
07:52:39  * dguttmanquit (Quit: dguttman)
07:57:54  <Raynos>rvagg: have you read chromium indexeddb implementation?
08:16:27  * AvianFluquit (Remote host closed the connection)
08:34:18  * No9joined
08:44:18  <mbalho>prototypes aren't hoisted!
08:58:53  <niftylettuce>anyone here from austin, tx?
09:04:39  * dominictarrjoined
09:27:57  * st_lukequit (Remote host closed the connection)
09:46:29  * jibayjoined
10:05:12  <dominictarr>heh, scuttlebutt is on the front page of google results for "scuttlebutt"
10:05:24  <dominictarr>as well as a bunch of sailing stuff
10:53:05  * fotoverite_joined
10:54:11  * fotoveritequit (Ping timeout: 255 seconds)
10:54:11  * fotoverite_changed nick to fotoverite
12:09:48  * shuaibjoined
12:21:19  * jibayquit (Quit: Leaving)
12:24:31  * dominictarrquit (Quit: dominictarr)
12:28:23  * dominictarrjoined
13:21:29  * AndChat|648704joined
13:21:36  * AndChat648704quit (Read error: Connection reset by peer)
13:21:57  * AndChat648704joined
13:25:34  * AndChat|648704quit (Ping timeout: 240 seconds)
13:55:34  * No9quit (Ping timeout: 240 seconds)
13:57:53  * No9joined
13:58:34  <dominictarr>Raynos: ping?
14:26:08  * No9quit (Ping timeout: 255 seconds)
14:27:47  * No9joined
14:46:11  <dominictarr>Raynos: sockjs-stream doesn't work for me from the server with reconnect … that is what it's for, right?
14:53:29  <ralphtheninja>http://www.youtube.com/watch?v=-RMOGFaOLSQ&feature=related
15:03:43  <dominictarr>the universe is obviously _at least_ a computer
15:03:52  <dominictarr>look at how a computer works,
15:04:24  <dominictarr>a computer is a logical machine that is sufficiently complicated that you can use it to simulate a computer.
15:04:57  <dominictarr>their is a ceiling on complexity, it's either a computer or less that a computer
15:05:27  <substack>fallacy of composition?
15:05:37  <dominictarr>take a computer, you can simulate another computer on it, then simulate another computer on that computer, etc etc
15:05:48  <dominictarr>computers all the way in
15:06:02  <dominictarr>VMs running inside VMs,
15:06:49  <dominictarr>so, if you can build a computer it in, the universe must be _at least_ a computer
15:07:22  <dominictarr>whether it's somehow more than a computer, superturing, that is another question.
15:07:54  <pkrumins>what are you smoking
15:07:57  <substack>s/computer/widget/
15:08:04  <substack>and it still holds
15:08:21  <substack>or bread
15:08:59  <substack>if you can bake bread in a universe the universe must be at least bread!
15:10:06  <guybrush_>or a bread oven
15:10:28  <dominictarr>no, but you can't simulate bread inside a bread
15:10:48  <substack>still, it seems somewhat suspect when computer programmers go around saying that the universe is akin to a giant computer
15:10:59  <dominictarr>you can't use a bread to calculate how any other possible bread would respond
15:12:00  <guybrush_>though i would rather use "machine" than "computer"
15:12:26  <dominictarr>a computer is just a special class of machine
15:12:33  <dominictarr>an information machine
15:13:12  <guybrush_>it cant be that everything is virtual, there has to be a beginning!
15:13:57  <substack>why?
15:14:14  <guybrush_>im not sure actually now that i think about it
15:14:17  <substack>our intuitions about what should or shouldn't be the case are suspect
15:14:52  <substack>they were built by evolution to deal with medium-scale local phenomena
15:15:14  <guybrush_>its hard to think about anything without a beginning
15:15:53  <guybrush_>"it just exists", doesnt work for me
15:15:55  <dominictarr>guybrush_: well, you get to a bit outside what we can ever reason about, because it's impossible to gather information
15:16:08  <dominictarr>imagine if you lived inside one of our computers...
15:16:13  <substack>being hard to think about doesn't preclude something from being false
15:16:15  <dominictarr>could you learn about RAM
15:16:17  <guybrush_>in the matrix!
15:16:21  <substack>s/false/true/
15:16:29  <dominictarr>or L1, L2 caches?
15:16:43  <dominictarr>you might be able to discover that experimentally
15:16:47  <guybrush_>right substack but than it wouldnt make sense to argue about anything
15:17:28  <dominictarr>the point of calling the universe a computer, is that it's a model
15:17:34  <substack>I'm not making an argument for sophistry
15:17:50  <substack>dominictarr: I prefer to classify that as a metaphorical tool
15:18:03  <substack>it gives us the language to frame our thinking
15:18:07  <dominictarr>it's not the "truth" it's just about an idea and how well it fits evidence
15:18:17  <dominictarr>yes, that is all we have
15:18:30  <substack>but you've got to be careful not to ascribe attributes falsely
15:18:56  <substack>like for instance we build machines and computers but it would be a mistake to ascribe teleology
15:19:04  <substack>to the universe at large
15:20:13  <guybrush_>i have to google teleology
15:20:14  <dominictarr>people think of computers as electronic devices
15:20:22  <dominictarr>but that is incorrect.
15:20:31  <dominictarr>a computer is a logical construct.
15:21:03  <dominictarr>a set of constructs, rather
15:22:03  <guybrush_>i like to think that its more like chaos, i mean everything in the universe we live in is just chaos
15:22:10  <guybrush_>not logic
15:22:47  <dominictarr>ah, but chaos is actually a logical (mathematical) construct too!
15:22:59  <guybrush_>ha
15:23:02  <dominictarr>guybrush_: have you heard about how they discovered chaos?
15:23:18  <guybrush_>no but i am interested
15:23:39  <guybrush_>who discovered chaos haha
15:23:53  <dominictarr>Edward Lorentz
15:24:11  <dominictarr>" In 1961, Edward Lorentz discovered the butterfly effect while trying to forecast the weather. He was running a long series of computations on a computer when he decided he needed another run. Rather than do the entire run again, he decided to save some time by typing in some numbers from a previous run. Later, when he looked over the printout, he found an entirely new set of results. The results should have been the same as before. After thinking
15:24:11  <dominictarr>about this unexpected result, he discovered that the numbers he typed in had been slightly rounded off."
15:25:00  <dominictarr>the thing with chaos is that a very small change in initial conditions creates large scale changes after a sufficient period
15:25:22  <dominictarr>http://www.schuelers.com/ChaosPsyche/part_1_3.htm
15:26:00  <guybrush_>why the hell is it edward lorenz without "t" on wikipedia
15:26:43  <guybrush_>anyway interesting stuff
15:27:34  <dominictarr>yes, incidentally, turns out humans needed computers to discover chaos theory.
15:28:00  * AvianFlujoined
15:29:30  <dominictarr>gtg. catch you dudes later.
15:29:32  * dominictarrquit (Quit: dominictarr)
15:31:29  <ralphtheninja>chaos is also related to things not being determenistic
15:31:52  <guybrush_>quantum computers!
15:31:53  <ralphtheninja>deterministic*
15:32:26  <ralphtheninja>I agree that the universe must be at least a computer
15:33:06  <guybrush_>it depends on the definition of "computer" :D
15:33:12  <ralphtheninja>at least a turing machine is a more appropriate description :)
15:37:03  <pkrumins>universe is just a regular expression
15:41:22  * jibayjoined
16:31:45  <isaacs>anyone in here use websockets for stuff? (and not socketio)
16:31:50  <isaacs>Raynos, substack ^?
16:31:54  <isaacs>mbalho: ^?
16:41:15  * substackdoes
16:41:39  <substack>by way of shoe which uses sockjs
16:44:21  <isaacs>substack: so.. when you get the upgrade event, it just grabs the actual TCP socket and uses that, right?
16:44:44  <substack>I think so
16:45:49  <substack>are you poking at new http api ideas?
16:46:29  <substack>ideal websockets api: if (req.headers.upgrade) req.pipe(myStream).pipe(res)
16:46:50  <substack>where `myStream` would handle the decoding or whichever
16:47:17  * mikealquit (Quit: Leaving.)
16:54:38  <isaacs>substack: not quite yet.
16:54:45  <isaacs>i'm poking at not breaking the existing http api
16:54:53  <isaacs>which i'm pretty sure i did in 0.9 with streams2
16:54:56  <isaacs>but i'd like to fix
16:56:11  <isaacs>ralphtheninja: actually, the butterfly effect that Lorentz talked about *depended on* the universe being deterministic
16:56:32  <isaacs>ralphtheninja: without being deterministic, there's no "sensitive dependence on initial conditions", because there's no "dependence on initial conditions"
17:01:46  * tilgovijoined
17:02:02  * saijanai_joined
17:02:57  * shuaibquit (Ping timeout: 276 seconds)
17:08:26  * sorensen_joined
17:10:33  * mikealjoined
17:15:50  * yorickjoined
17:17:49  * dguttmanjoined
17:26:29  * No9quit (Ping timeout: 244 seconds)
17:27:54  * No9joined
17:28:47  * jibayquit (Remote host closed the connection)
17:36:36  <Raynos>isaacs: I use sockjs / engine.io
17:48:37  <ralphtheninja>isaacs: ok, so the chaos is generated by a deterministic universe
17:49:17  * dominictarrjoined
17:50:12  <ralphtheninja>it makes sense, just like cellular automatas can generate seemingly chaotic behavior
18:01:59  * AndChat648704quit (*.net *.split)
18:01:59  * owen1quit (*.net *.split)
18:02:27  * AndChat648704joined
18:02:42  * owen1joined
18:03:16  * shuaib_joined
18:04:38  * st_lukejoined
18:04:55  * shuaib_changed nick to shuaib
18:06:25  * tilgoviquit (Remote host closed the connection)
18:10:38  * dominictarrquit (Quit: dominictarr)
18:16:36  <isaacs>ralphtheninja: of course, strict "determinism" is not necessarily for observable deterministic effects. only a certain level of statistical predictability.
18:17:11  <isaacs>but SDIC depends on some dependence on initial conditions. truly random systems don't exhibit SDIC
18:20:34  <mbalho>isaacs: i agree that the ideal websockets api is if (req.headers.upgrade) req.pipe(myStream).pipe(res)
18:20:49  <CoverSlide>unfortunately a computer can only make measurements in integral units, and non-integral measurements are merely emulated. This goes for both spatial and temporal measurements. As of right now, we know of no distinct atomic unit of time or space, so most likely the universe in not a computer
18:21:18  <mbalho>isaacs: i use the ws module for hosting websocket connections e.g. https://github.com/maxogden/voxel-server/blob/master/server.js#L35-L36
18:21:52  <mbalho>isaacs: but it would be way awesome if there was a good way to attach a websocket server to an existing http server e.g. what substack suggested
18:22:13  <mbalho>isaacs: cause right now you have to remove all listeners and then add them to your own proxy, effectively monkeypatching an existing server
18:23:25  <substack>please break backwards compat on that
18:23:48  <substack>the way core http works now is encouraging so many painful interfaces
18:24:32  <CoverSlide>remove http from core
18:25:00  <CoverSlide>needs to be done at some point
18:38:05  <isaacs>substack: we wont' break back compat in 0.10. but we will in 0.12, probably
18:38:17  <isaacs>substack: especially if loudmouth community people call for that breakage.
18:38:37  <isaacs>substack: (ie, you, mbalho, Raynos, etc. be loudmouths. call for breakage.)
18:39:49  <fotoverite>isaacs: true, a random system could look totally deterministic though. :P It's just randomly doing that.
19:02:08  * wiwilliajoined
19:02:56  <wiwillia>Hey pkrumins, any chance you have experience with Amazon s3?
19:04:48  <CoverSlide>with enough variables, anything can look random
19:05:52  <CoverSlide>a coin flip is deterministic, based on weight of the coin, force and aposition of the thumb, the placement of air molecules, the properties of the surface it lands on. the variables are immeasurable
19:06:04  <CoverSlide>random is just another word for "too many variables"
19:09:22  <mbalho>isaacs: in my defense im not a loudmouth usually
19:09:39  <mbalho>mbalho: i usually dont care either way but the http websocket thing is just ugly
19:12:54  <fotoverite>mbalho: Loudmouth no but you are opinionated
19:14:54  * dguttmanquit (Quit: dguttman)
19:18:49  <mbalho>fotoverite: i am but a humble member of the Church Of Node Core
19:20:47  <fotoverite>I don't think I can make nodeconf this year, :(
19:20:51  <fotoverite>guess summer camp.
19:21:13  <fotoverite>people are doing good work for nodepdx
19:23:44  <ralphtheninja>CoverSlide: agree, but you don't necessarily need many variables or complex algorithms to create a random behavior
19:24:07  <CoverSlide>in other words, pseudo-randomness
19:24:50  * tilgovijoined
19:27:07  <ralphtheninja>yeah I guess
19:29:34  <ralphtheninja>the process of flipping a coin might have many variables, but even that is pseudo random in that case .. just have more variables
19:31:37  <ralphtheninja>which in turn means that 100% randomness can never be achieved .. it's just different orders of complexity .. different degrees of randomness .. or?
19:32:10  <isaacs>CoverSlide: "random" literally means "impossible to predict"
19:32:15  <isaacs>not "without a cause"
19:32:48  <isaacs>the difference between "pseudorandom" and "random" in computer lingo is that, if you know the seed, you CAN predict pseudorandom variables.
19:33:01  <ralphtheninja>check
19:33:03  <isaacs>(er, know the seed and the algo)
19:33:20  <isaacs>random = pseudorandom algo seeded with machine entropy
19:33:32  <isaacs>mouse movement, etc.
19:33:45  <ralphtheninja>which is the same thing as a cellular automata .. you can reproduce the output 100% every time .. but the patterns might still be seemingly random
19:33:47  <isaacs>fan speeds, CPU heat, other things that are hard to guess
19:33:53  <isaacs>right
19:34:07  <CoverSlide>which brings to mind this article which surfaced on reddit recently: http://www.empiricalzeal.com/2012/12/21/what-does-randomness-look-like/
19:34:29  <ralphtheninja>a night of randomness :)
19:34:31  <CoverSlide>Which talks about Poisson distributions
19:34:53  <isaacs>unless you reject the many-worlds interpretation (which is silly to do) there's nothing that is so random as to be causally disconnected from a prior state.
19:35:06  <isaacs>but it IS random (as in, unpredictable) which universe you'll end up in sometimes.
19:36:24  <rowbit>SubStack, pkrumins: At least 5 people waiting in the queue for free servers! (Waiting: 15)
19:38:33  <CoverSlide>free servers?
19:44:56  * dguttmanjoined
19:46:50  <Raynos>isaacs: what are we breaking?
19:47:33  <CoverSlide>http
19:51:07  * sorensen_quit (Quit: Bye!)
19:53:00  <mbalho>CoverSlide: speaking or PRNGs i use your alea module to seed this http://maxogden.github.com/voxel-simplex-terrain/?seed=mattdamon
19:53:29  <CoverSlide>awesome
19:53:57  <st_luke>substack: I would argue that most caching in web apps nowadays isn't for saving bandwidth but for reducing db queries
19:54:04  <st_luke>just from what I've experienced in my tenure in the hosting industry
19:54:42  <CoverSlide>yeah the algo isn't mine, but I added some stuff specifically for syncing between client and server
19:54:58  <substack>https://github.com/substack/gray-code
19:55:01  <CoverSlide>haven't used it for anything as of yet
19:55:19  <substack>isaacs: sooner or later every algorithm with a wikipedia entry will have its own npm module
19:55:20  <mbalho>substack: wat
19:55:27  <CoverSlide>haha
19:55:33  <mbalho>CoverSlide: i found it in github.com/jwagner/voxelworlds/ actually
19:56:24  <mbalho>isaacs: also every algorithm on npm will be able to be visualized in voxels
19:56:45  <mbalho>substack: spun this out https://github.com/maxogden/player-physics
19:56:48  <substack>mbalho: oh the default with voxel-debris is 1, more impressive when it's >= 4
19:57:16  <substack>yield that is
19:57:53  <substack>var explode = require('voxel-debris')({ yield: function () { return 4 } })
19:58:02  <substack>var explode = require('voxel-debris')(game, { yield: function () { return 4 } })
19:58:26  <mbalho>sweet
20:01:08  <mbalho>substack: pushed 0.2.1 of voxel-engine -- has some potentially breaking internal apis
20:01:58  <mbalho>substack: basically ive been moving all dom + rendering code into functions and all the state and physics into other functions so the engine can run in node and compute physics without rendering anything to the screen
20:02:04  <mbalho>substack: but i think thats done now
20:02:49  <substack>sweet!
20:03:30  <substack>although everything should be able to run entirely in the browser too
20:03:56  <mbalho>yes
20:03:56  <substack>but the freedom to move the computation around is certainly a plus
20:04:33  <substack>we can test all that code independently too
20:04:56  <mbalho>substack: last night i got my multiplayer server to run the engine and generate voxels in memory from a seed and then when clients join they get the game config sent to them and they generate the same base world
20:07:55  <CoverSlide>that's awesome, no need to send the full world data over the wire
20:08:17  <mbalho>CoverSlide: yea and then i just have to sync edits which is a few lines thanks to scuttlebutt
20:08:28  <mbalho>CoverSlide: hard part is what im about to dive into which is syncing the game loop on both sides
20:12:43  <substack>mbalho: will I be able to game.createStream() ?
20:17:32  <mbalho>substack: what would it be a stream of
20:18:31  <CoverSlide>state modifying events?
20:20:13  <substack>mbalho: a serialization of the game update data that the other clients will need in order to synchronize the game state
20:22:26  <Raynos>substack: it might be cleaner to have state seperate from game
20:22:43  <Raynos>otherwise game object becomes god object
20:23:05  <Raynos>game(state); stream.pipe(state.createStream()).pipe(stream)
20:23:21  <substack>game can just make a state object itself
20:23:52  <substack>the game has state already that would need to be serialized
20:35:18  * shuaibquit (Quit: Textual IRC Client: http://www.textualapp.com/)
20:45:45  <mbalho>substack: eventually yes i would like the game to be a stream
21:00:44  <jez0990_>on the topic of serialisation - is mux-demuxing other mux-demux streams (potentially recursively) likely to have any limitations?
21:03:03  <hij1nx>oh man, there's nothing interesting going on in the node room anymore...
21:12:38  <substack>it's just a help channel
21:14:33  <CoverSlide>time to move on to something else
21:14:36  <CoverSlide>go here i come
21:16:00  <fotoverite>Isn't that why we all stay in stackvm?
21:18:08  <CoverSlide>#Node.js - newbie room #libuv - node dev room #stackvm - cool hacker room
21:19:39  <chrisdickinson>cool guy shades are handed out at the door.
21:29:29  <isaacs>mbalho, substack: Yes! put every algorithm in npm! visualize every npm in voxels! put every voxel visualization algo in npm and visialize in voxels! yo dawg!!
21:32:11  * dguttmanquit (Quit: dguttman)
21:35:09  * AndChat648704quit (Read error: Connection reset by peer)
21:47:11  <Raynos>jez0990_: its inefficient
21:50:23  * jibayjoined
21:51:29  <mbalho>substack: do you have an illustrated self portait or do you prefer http://substack.net/images/substack_240.png
21:52:03  <mbalho>substack: also you should render this in voxel http://substack.net/images/substack_pixel.svg
21:52:10  <CoverSlide>That looks like professional photography to me
21:52:34  <jez0990_>Raynos: definitely agree with that sentiment, but I don't think there's any alternative. The module will have one of the mux streams being a control/registry stream that defines new demux streams to watch out for as they're generated
21:52:36  <substack>mbalho: I like the robot avatar the best
21:52:40  <substack>agreed @ voxelization
21:52:41  <mbalho>substack: kewl
21:53:07  <substack>there's also http://substack.net/images/avatars/robot_avatar.png
21:54:15  <mbalho>oh sweet
21:57:33  * mikealquit (Quit: Leaving.)
21:58:21  * mikealjoined
21:59:29  * CoverSlidequit (Quit: REBOOTIN')
22:03:10  * CoverSlidejoined
22:08:33  * tphummeljoined
22:16:00  <Raynos>jez0990_: the correct thing to do is to have a central muxxer and namespace the rest through it.
22:16:16  <Raynos>its the same with p2p networking
22:16:19  <Raynos>when building p2p apps
22:16:27  <Raynos>you want all your apps to use the same network bus
22:16:33  <Raynos>to the entire p2p network
22:16:48  <Raynos>and each app builds it own sub network on top of the low level network infrastructure
22:16:54  <Raynos>Now for sugar we can patch mux-demux
22:17:01  <Raynos>so that it flattens multiple muxes on muxes
22:17:33  <Raynos>mbalho: I want http://substack.net/images/substack_pixel.svg in a 3d voxel demo
22:17:45  <Raynos>I also want to be able to push it over and watch substack "die"
22:18:49  <mbalho>lol
22:20:58  <substack>blarg I am so tired right now >_<
22:21:36  <substack>need to finish this ui module to roll out some browserling pricing element updates
22:21:58  <substack>for team plans and then later charging for testling-ci
22:22:56  <substack>gotta keep that cash money rolling in
22:23:20  <mbalho>substack "bling bling" halliday
22:24:09  <substack>burrito time
22:29:16  <fotoverite>Agreed burrito time.
22:45:50  <jez0990_>Raynos: so I was only planning to use this nested mux stream as a oneway api exclusively between two different modules, but I think you're right about using the namespacing - thank you!
22:51:58  <pkrumins>defunctzombie_zz, Raynos, i just put sandbox browser isolation for testling-ci in production, now all the browsers run the right engines.
22:56:33  * devaholicquit (Ping timeout: 276 seconds)
23:02:25  * AvianFluquit (Remote host closed the connection)
23:02:46  * nk109joined
23:09:47  <mbalho>substack: we should render this in voxels https://github.com/Platane/Procedural-Flower
23:21:00  * devaholicjoined
23:28:04  <Raynos>jez0990_: I have a similar problem with https://github.com/Raynos/signal-channel/blob/master/connection.js#L14 and https://github.com/Raynos/peer-connection-shim/blob/master/open.js#L24
23:28:17  <Raynos>I build peer connection shim on top of a stream coming from signal channel
23:28:23  <Raynos>it just turns out that both want multiplexing logic
23:28:28  <Raynos>I think mux-demux should flatten it out
23:28:37  <Raynos>pkrumins: sweet!
23:30:36  * dguttmanjoined
23:35:04  <Raynos>substack: what do you think of HTML macros? ( https://gist.github.com/25cadf7725af6105e77a ) as a compile time preprocessor.
23:35:48  <Raynos>anyone else feedback welcome too :p
23:40:08  <kanzure>what
23:40:11  <kanzure>hrm.
23:49:14  <jez0990_>Raynos: that looks pretty logical to me, but I don't have any perspective on how or why it's unique compared with other forms of templating already out there :3
23:49:34  <Raynos>jez0990_: because logic is compile-time only and not run-time
23:50:32  * yorickquit (Remote host closed the connection)
23:50:37  <jez0990_>...and that would presumably "just work" with browserify?
23:52:18  <Raynos>yes / no
23:52:30  <Raynos>you have to tell browserify to use a tool to register these
23:52:44  <Raynos>but you would be able to compile them to a javascript file that works with browserify
23:53:15  <Raynos>or do `bundle.register(".template", require("macro-html")) `
23:54:17  <jez0990_>ah got it, cool
23:55:47  * AvianFlujoined