The Nineteenth Byte

Easterly Irk

22:00

@Quill try jQuery ad here

quartata

@Optimizer Uhhh because C++ is strict typed and you can't ever do something as inane as [] + ![]?

Optimizer

@quartata yeah, but still 'C' + 1 == 'D'

Easterly Irk

and get "false" as a answer. ;P

@Optimizer that actually makes sense though

quartata

@Optimizer Try it with "C" + 1

'C' is a char.

"C" is a char[].

Optimizer

I know

quartata

22:01

That makes perfect sense.

chars are numbers not strings.

Dennis

C doesn't have a character type.

Optimizer

all I am saying is that everyone has flaws

Alex A.

'C' + 1 == 'D' works in most if not all languages that distinguish between chars and strings

Dennis

It's called char, but it's just 8-bit integers.

quartata

Is it not 16-bit?

I have no idea.

Dennis

22:01

@quartata No.

Alex A.

@quartata Nope

Quill

you're literally incrementing ascii integers, of course it would work

Doorknob

@quartata char is 8 bits on all sane systems

quartata

@Dennis Oh

Optimizer

because typed variables are better

quartata

22:02

@Optimizer No.

Weak typing is fine

when it's done sanely

Dennis

Technically, it's 1 byte, whatever that means.

Optimizer

use TypeScript then

quartata

@Dennis 1 octet

@Optimizer That's my point. People are turning to other alternatives.

Optimizer

isn't Dynamic typing JS?

Doorknob

@QPaysTaxes technically not always

quartata

22:02

But they're not gaining traction

Dan Pantry

@Optimizer don't - TypeScript is bad at interop. Use flowtype

quartata

@Optimizer The problem is JS's aggressive type coercion.

Optimizer

@quartata its still JS though.

quartata

@Optimizer Only because it has to be

You have no choice but to transpile to JS because it's the only thing browsers support

Dennis

24

A: What platforms have something other than 8-bit char?

Machines with 36-bit architectures have 9-bit bytes. According to Wikipedia, machines with 36-bit architectures include: Digital Equipment Corporation PDP-6/10 IBM 701/704/709/7090/7094 UNIVAC 1103/1103A/1105/1100/2200,

quartata

22:04

Yeah, {} + [] === 0 is basically what I'm talking about

Optimizer

so life would be better if all browsers ever supported more than one language?

like JS and C# and Python?

quartata

@Optimizer Yes.

Optimizer

oh please

quartata

@Optimizer What's so unreasonable about that??

Quill

@EasterlyIrk da

Easterly Irk

22:04

Let's stahp argueing and go play tf2

wtf who removed my caret?

quartata

I'd really rather not have to use a language that does {} + [] === 0

Or even worse [] + {} === {}

Dan Pantry

in other news you could stop complaining about type coercion and just use strict mode.

Doorknob

@EasterlyIrk nothing of value was lost

6

Dan Pantry

like every serious javascript developer does.

Easterly Irk

@Doorknob D:

Dan Pantry

22:05

> (function() {
... 'use strict'
... console.log({} + [] === 0)
... }())
false

Easterly Irk

CARETS ARE LIFE

quartata

@DanPantry That's an ES6 feature right?

Optimizer

@quartata standard committee for each one, X times development effort in browsers, no new browser ideas seeing the daylight just because not enough manpower to support all languages and eventually you are left with just Google Chrome.

Dan Pantry

@quartata No. Strict mode has been around since at least ES5.

quartata

@Optimizer I'm saying that if it could happen it would be great, not that it will.

Dan Pantry

22:06

Everyone knows that non-strict mode sucks.

Quill

I thought strict mode was like ES3 or 4

Dan Pantry

That's why everyone uses strict mode.

quartata

@DanPantry Some people don't apparently

Although that's good to know

Easterly Irk

@Doorknob stahp pls.

quartata

I'm still pissed that you even have to do something like (function() { ... })() for a new scope

Optimizer

22:07

@quartata I am saying that how do you know that it will be great? from what I see its going to end like what I just wrote.

Easterly Irk

@QPaysTaxes [Object object]

Doorknob

@EasterlyIrk ... I didn't touch anything

Optimizer

already browsers are finding it hard to cope up with just 1 standard

Easterly Irk

Who deleted my carets?

@Doorknob Sorry for accusing you.

Dan Pantry

@QPaysTaxes "[object Object]" because yes, it casts it to a string, even in strict mode. I concede that is bad.. but how often do you find yourself adding an object to an array? Do you have any real-life examples of that happening?

Easterly Irk

22:07

If you really don't like carets, just tell me?

Dennis

@EasterlyIrk That would have been me. They add nothing to the conversation.

Optimizer

^^

Doorknob

@QPaysTaxes ( ͡° ͜ʖ... okay, wait, why am I Lennying myself

9

Easterly Irk

Okay.

Quill

Browsers are finding it hard to stick to a standard because they're all racing each other for dominance

Dan Pantry

22:08

@QPaysTaxes Did you only read the fact that I said the result and ignore the other 3/4s of what I said/asked?

quartata

@DanPantry What if you typo something and end up with the wrong variables?

The error occurs completely silently

Dan Pantry

@quartata That's why you use static analysis, or unit tests

@QPaysTaxes My apologies

quartata

@DanPantry You shouldn't have to. Every other language would warn you.

Infinitely less effort.

Dan Pantry

@quartata Every other language warns you because of their type system. JavaScript doesn't have that.

Optimizer

@Quill which is both good and bad, otherwise there's no motivation to develop. stagnant market. look at how much time java 8 took, and 9 will.

Dan Pantry

22:09

Ultimately, JavaScript's type system can be useful but gives you more rope to hang yourself with in that specific case.

quartata

@DanPantry False. Any dynamic typed language under the sun would warn you

Python, Ruby, etc...

Quill

Unless you change the definition for the + operator, you can't throw errors on type comparisons or concatenation. Even then, it'll still go to primitive

Dan Pantry

So.. has anyone got a realistic example of aggressive coercion in strict mode that isn't {} + [] :P

quartata

@QPaysTaxes oops pretend you didn't see that

Cᴏɴᴏʀ O'Bʀɪᴇɴ

\o/ my internet's back on!

Dan Pantry

22:10

and that can't be easily prevented by simply using Flowtype (which comes for free with Babel - you are using Babel ,right?) or common sense

Easterly Irk

@CᴏɴᴏʀO'Bʀɪᴇɴ \o/

HAI

Dan Pantry

wb, @CᴏɴᴏʀO'Bʀɪᴇɴ

Easterly Irk

Did you see my jolf message?

Cᴏɴᴏʀ O'Bʀɪᴇɴ

@DanPantry Hi!

@EasterlyIrk Yes!

quartata

No because why should I bother to use a language where I have to use a third-party tool just to make it work sanely?

The only reason why is because I have to.

Optimizer

22:10

@DanPantry so its equivalent of me using IntelliJ on normal JS?

quartata

/end point

Easterly Irk

@CᴏɴᴏʀO'Bʀɪᴇɴ would work on it now but rewriting essays.... Q_Q

Cᴏɴᴏʀ O'Bʀɪᴇɴ

:(

Easterly Irk

ikr

Dan Pantry

@QPaysTaxes @quartata Well, I am a JavaScript developer and honestly I've never once had to deal with a type coercion issue. Obviously, anecdotes are anecdotes, but that's my 2c

Easterly Irk

22:11

welp, bai to actualy get work done

Dan Pantry

@QPaysTaxes Babel is not huge. And you're using Babel primarily for backwards compatibiltiy (not all browesrs support es6 yet) - you just get flowtype thrown in for free.

Optimizer

@quartata why don't you start writing in ASM? it's pure JS, just instead of "use strict", use "use asm"

Quill

If we want to understand the type system's internal logic a bit better: 418stat.us/2016/04/javascripts-type-system

Dan Pantry

@QPaysTaxes I have found it very difficult to use rigid type systems when you want to use union or intersection types, which are surprisingly quite common in js

for example, mixins are not so easy to do in a strong type system unless they are explicitly planned for (scala)

Quill

the caret thief strikes again!

Dan Pantry

22:13

s/static/strong

Cᴏɴᴏʀ O'Bʀɪᴇɴ

wtf

Dan Pantry

> unless they are explicitly planned for

Cᴏɴᴏʀ O'Bʀɪᴇɴ

who deleted my message

Doorknob

8 mins ago, by Doorknob

@EasterlyIrk nothing of value was lost

Easterly Irk

@CᴏɴᴏʀO'Bʀɪᴇɴ Dennis.

Dan Pantry

22:14

I was also mainly referring to C# where the closest you get to a mixin is an Attribute

Easterly Irk

He doesn't like carets.

Dan Pantry

carets don't add to the discussion

Easterly Irk

6 mins ago, by Dennis

@EasterlyIrk That would have been me. They add nothing to the conversation.

Doorknob

@EasterlyIrk Actually, he didn't

Cᴏɴᴏʀ O'Bʀɪᴇɴ

wtf

Easterly Irk

22:14

It's fine.

Cᴏɴᴏʀ O'Bʀɪᴇɴ

I strongly protest that.

Easterly Irk

eh

not argue right now

go yell on skype

Alex A.

9 mins ago, by Doorknob

@EasterlyIrk nothing of value was lost

Easterly Irk

^ He has a point, and he is a mod. Deal with it.

Cᴏɴᴏʀ O'Bʀɪᴇɴ

That's like saying "I agree" is a pointless thing to say in a conversation.

Optimizer

22:15

so... anything .. beside type system?

Dan Pantry

I mean, there's a lot of hate on JS but I find it quite a nice language to work with once you step around the potholes (loose mode, double equals, don't pass value types to .call or .apply). Outside of that I find the flexibility quite useful. And it lends itself well to structural - rather than nominal - typing :-)

Quill

Interesting to note that the deletion of noise simply creates more noise

Dan Pantry

And to be clear, there are potholes in JS.

Cᴏɴᴏʀ O'Bʀɪᴇɴ

@Quill I concur strongly with this point.

Dan Pantry

@QPaysTaxes Because 'use strict'; is an expression that on it's own does nothing. It'll have no side effects.

Meaning that browsers that don't iunderstand it will simply ignore it.

Optimizer

22:16

@QPaysTaxes oh, now you have problems with choices? :P

(backward compatibility)

Cᴏɴᴏʀ O'Bʀɪᴇɴ

I further state that it's absolutely absurd to micro manage peoples chat messages for what you claim to be noise.

Easterly Irk

Lets go back to talking about golfing ~~and avocad~~.

If we ever did.

/s

Optimizer

its juic avocad not golf avocad

Cᴏɴᴏʀ O'Bʀɪᴇɴ

Where's the line? Half of these messages can be classified as noise.

Dan Pantry

JavaScript is a language built on bandaids and backwards compat. ES6 is actual the tipping point where standards agencies have gone "Nah, we're tired of this crap now. Let's actually make ourselves viable". Though they still aren't getting rid of double equals. @_@

@QPaysTaxes Double equals is loose equality.. sort of

Cᴏɴᴏʀ O'Bʀɪᴇɴ

22:18

Is this not noise?

1 min ago, 25 seconds total – 6 messages, 3 users, 0 stars

Bookmarked 5 secs ago by Cᴏɴᴏʀ O'Bʀɪᴇɴ

Easterly Irk

woah was that lag or lots of messages at once

@CᴏɴᴏʀO'Bʀɪᴇɴ @CᴏɴᴏʀO'Bʀɪᴇɴ you won't win. I am sorry.

Dan Pantry

@QPaysTaxes '5' == 5 is true. '5' === 5 is false (type coercion)

Easterly Irk

@QPaysTaxes huh

Quill

@QPaysTaxes type coerceon... there's a long table of effects on SO if you like

Cᴏɴᴏʀ O'Bʀɪᴇɴ

@EasterlyIrk I can still argue the point.

Easterly Irk

22:19

bai finding more interesting chat

Quill

3761

A: Does it matter which equals operator (== vs ===) I use in JavaScript comparisons?

The identity (===) operator behaves identically to the equality (==) operator except no type conversion is done, and the types must be the same to be considered equal. Reference: Javascript Tutorial: Comparison Operators The == operator will compare for equality after doing any necessary type c...

Dan Pantry

@QPaysTaxes well, when you think about it sort-of kind-of makes sense.. back when it was made, every attribute on a DOM el was a string. It just meant if you had an attribute that was numeric you wouldn't need to parseInt it.

Cᴏɴᴏʀ O'Bʀɪᴇɴ

@QPaysTaxes constant. That was one.

Dan Pantry

But it has a whole slew of side effects; every static analyser and linter and strict mode will yet at you if you try and use it

Optimizer

so does that mean that we cannot start a caret party?

the meme is obsolete ????

Dan Pantry

22:20

@QPaysTaxes Sure, but the difference is that C99 was a version of compiler, and by and large the byte code generated by it is still usable, right?

But you can't do that in JS, because it's interpreted and not compiled (sort of)

Optimizer

@QPaysTaxes you also need to understand that upgrading to C99 was a choice.

Dan Pantry

Until there are about.. 0.01% of the user base using ES5, then we will still have to compile down to ES5.

You can't convince users to upgrade their browsers. It doesn't matter for C++ where you can statically link all the libraries and just deploy that on it's own with no 3rd party dependencies

Easterly Irk

Any more answers here? The bounty was meant to attract more than 1.

Dan Pantry

Also bear in mind that every user who can't use your site is a potential lost customer

Optimizer

@DanPantry i think the blocking point is not that users dont upgrade browsers, but that site developers don't want to upgrade their broken sites in the upgraded browsers

Dan Pantry

22:22

@Optimizer eh? I don't follow

In my most recent project we were forced to support IE9 because the business wanted it until IE9's perf issues forced them to upgrade to a more modern browser - we wanted to code it in modern standards with flexbox and such but couldn't until that tipping point

Optimizer

backward compatibility in standard is required to make sure older sites still keep working.

Dan Pantry

@Optimizer That's true, but the backwards compatibility is only required because of older browsers

If everyone tomorrow upgraded to Edge, Chrome 48, whatever, then there'd be no reason to keep the IE9 compatibility code in

Optimizer

not browsers.. but sites..

Dan Pantry

@Optimizer Maybe I'm misunderstanding. From my POV, the code I write now in ES6 is not going to impact geocities for example. But it's not going to work in someone's browser who is more than a year old.

Optimizer

@DanPantry so if all these only supported ES5+ then hell lot of sites will break. and that is a more important concern than people still being stuck to IE9

Dan Pantry

22:24

@Optimizer Ah. I see what you mean.

@QPaysTaxes To be fair, a large reason why our business personally was so wanting to stick to IE9 was security concerns.

IE9 is a lot more.. configurable for enterprise solutions than Chrome is.

Mars Ultor

But can't you mostly use polyfills/shims?

Dan Pantry

@QPaysTaxes Uh, not from external attacks. Internal ones.

Optimizer

@DanPantry but not more than IE 10 or Edge, right?

Dan Pantry

So with IE it's a lot easier to stop users going to a certain site. With Chrome, we had to redirect all users to a proxy that only allows specific whitelisted IPs through.

Optimizer

@DanPantry installing a DNS layer blocker app is not that hard. our company does that.

Dan Pantry

22:26

@Optimizer Well, IE10 and IE11 both also were choking on AngularJS. Turns out IE10/11 have some kind of setting turned on by default that causes it to choke.

Also it was AngularJS 1.3 which is really bad for performance, but that's not my choice of tech

Quill

TIL: you can write a custom action for ToPrimitive: gist.github.com/The-Quill/6d31d4aef337360fdd2855c89d8af546

Dan Pantry

All of our internal systems are still on W7, so no Edge

@QPaysTaxes @Optimizer I didn't make the infrastructure calls, I'm just explaining in laymans what was told to me by our itsec

quartata

@Quill Interesting

Dan Pantry

Also, "it's just an internal application" ergo they don't want to spend too much resource on it..

Quill

I wonder whether I could overload the Object constructor to apply that value to every incoming object

Optimizer

22:28

@Quill that's news even for me..

Dan Pantry

@QPaysTaxes Well, yes and no. Basically, the whitelist was only to be applied to outbound network connections from a virtual machine farm. So I am pretty sure they could have done it, but it was a matter of time and resource - they only had 2 days to set this up because we deployed the app in production to IE11 to 200 users, replacing their old desktop app (which was broken due to schema changes in the new app).

quartata

@QPaysTaxes That would be a good thing

It would mean he could override how type coercion works for everything

i.e he could make it good

Dan Pantry

So it was damage control to get Chrome 48 up and running ASAFP. I guess the whitelist with IPs over a proxy on the virtual machine farm was easier.

That was a really stressful week.

Yes, and also the gain would be pretty minimal.

Quill

It would change the basic spec definition for ToPrimitive (for variables of that type)

Dan Pantry

The users of that app are internal users and would only be using Chrome for that app anyway.

Optimizer

22:31

@Quill ah, its ES7

Dan Pantry

@Optimizer s/7/2016

Quill

Yeah, I know...

Dan Pantry

(apparently versions are being named after years now)

Optimizer

@DanPantry it'll still remain ES7/8/9/10 in my heart

Dan Pantry

@Optimizer please tell me there isn't an es8/9/10 yet. i barely know all the features in 7

what is there.. async/await, decorators, bind syntax..?

Optimizer

22:32

@DanPantry why do you want to? you dont even know if they'll stay in.

Quill

array deconstruction

Dan Pantry

i believe array comprehensions got scrapped

Optimizer

what is decorator?

Dan Pantry

@Quill array destructors are in es6

Optimizer

@DanPantry that was from es6. its there in 7

Dan Pantry

22:32

@Optimizer async functions are definitely staying

bind syntax was strawmanned though, and it is a bit weird anyway. ::this.item === this.item.bind(this)

@Optimizer Decorators look like this

Optimizer

yeah, remembered after a quick search

Dan Pantry

class MyComponent {
  @Inject foo;
}

Optimizer

^ and people say its not Java syntax

Dan Pantry

They're definitely staying in because angular and also TypeScript has already adopted them, but apparently the spec changes more often than Hilary Clinton's political views

Which is why Babel currently has them disabled

Easterly Irk

> Hillary Clinton's political views

Optimizer

22:35

but unless they make some default decorators, its not much of use

Easterly Irk

can i use that? :P

Dan Pantry

@Optimizer They're cool, but I'm not really seeing a huge use for them. Additionally, the current shims for them are very bulky, which makes me not want to use Angular 2 (as Angular 2 requires them).

bit-political-in-here

@EasterlyIrk sure

Optimizer

@EasterlyIrk caniuse.com should tell you the answer

Dan Pantry

@Optimizer The main thing is decorators are very useful with the current component-based development pattern.

Optimizer

@DanPantry but not if you have to write all decorators yourself and then people just keep duplicating things

Easterly Irk

22:36

@Optimizer apparently not

caniuse.com/#search=hillary%20clintons%20political%20views

Dan Pantry

import connect from 'react-redux'

@connect(state => ({ greeting: state.greeting }))
const GreetingComponent = ({ greeting }) => <h1>{greeting}</h1>

export default GreetingComponent

(bad example as it's ugly, but that's what it is used for)

quartata

@DanPantry What is this?

I mean what language

Dan Pantry

@quartata JS

Optimizer

I hate when strings don't need quotes

in Angular, react and JSX

Dan Pantry

Though I might have gotten the syntax.. slightly wrong

Easterly Irk

22:38

The universe started in 1970. Anyone claiming to be over 38 is lying about their age.

Dan Pantry

@Optimizer after using Angular for a year and seeing react, I hate Angular. The way you pass props through the component stack is messy as all hell

Mars Ultor

@Optimizer And JSX?

Easterly Irk

rofl

Dan Pantry

Yeah sorry I should mention what I have there is ES6 + JSX

Optimizer

@DanPantry what is JSX?

Dan Pantry

22:39

@Optimizer I have no clue what the acronym stands for but it is basically syntactic sugar around React.createElement calls

Optimizer

ok

Easterly Irk

@QPaysTaxes this one too

I learned from Achewood that since this poem is in ballad meter, it can be sung to the tune of Gilligan's Island. Since then, try as I might, I haven't ONCE been able to read it normally.

Dan Pantry

<h1>Hello, world</h1> becomes React.createElement('h1', {}, 'Hello, world!')

Optimizer

i hate it when you write your code as string and a JS code parses it without using eval

Mars Ultor

greeting => $('<h1/>').text(greeting) <- jQuer is the best

Dan Pantry

22:40

@MarsUltor no

Optimizer

@DanPantry it looks better

at least my strings have quotes

Dan Pantry

React.createElement('h1', {}, 'Hello, world!') for one is more like a Haskell IO Monad

it doesn't create an element like that jQuery snippet does

Quill

So I got a defined constructor method to do it: gist.github.com/The-Quill/…

Dan Pantry

@Optimizer since when do you write HTML code with quotes? :P

Quill

But I don't know how to override the default wrapper constructor yet

Optimizer

22:41

@DanPantry since when do you write HTML tags in JS?

Quill

since React. Also Hack (Facebook's PHP dialect) has a similar notation

Dan Pantry

@Optimizer Since writing nested React.createElement is hard to read

Mars Ultor

@DanPantry Since when do you do <div onclick=handleClick()></div>?

Dan Pantry

@MarsUltor We don't do that...

Well, I mean, we do and we don't

Optimizer

but it works

:D

Dan Pantry

22:42

It certainly looks like that, but it isn't an actual inline js event

Optimizer

lets not talk about HTML. even I agree that its a very bad language

Easterly Irk

let's talk about more noise stuf like avocad and avocad juic

Optimizer

no,JS was fine

Dan Pantry

well, crap.

Example of JSX before/after compile

Mars Ultor

JSX here

quartata

22:47

Wait, Babel has an online thing where you can transpile?

Dan Pantry

@quartata And evaluate. Its basically an online REPL

Quill

Like a tryitonline :D

Dan Pantry

you can also enable/disable the various presets and a few options like high compliancy

also stage-0 through stage-3 are experimental es7 features

stage-0 being the most experimental and stage-3 the most stable

quartata

From a file?

Dan Pantry

es2015-loose is es2015 but with a bit more "yo momma"

Luis Mendo

22:49

@quartata Hm. That sounds interesting, but if the matrix is square it gives "inconsistent" behaviour so to speak? I mean, if you want det(A^T * A) and A happens to be square...

Quill

@DanPantry That's actually what MDN says about it as well. good description

quartata

@LuisMendo True.

Mars Ultor

> Invalid preset specified in Babel options: "on"

quartata

@QPaysTaxes Why not just use a while loop? Read a character and break when the regex matches

@QPaysTaxes yes

Dan Pantry

@MarsUltor wtf did you do? lol

Mars Ultor

22:50

Nothing

Optimizer

@Quill so when you say that wish you could change object's contructor, you will still need to create that object using that constructor, right?

Mars Ultor

just tried to enable React compilation

Optimizer

like var a = new Object({a:4}) instead of var a = {a:4}

Quill

yeah, I can't overload the initializer declaration for the object type, I don't think

Dan Pantry

@MarsUltor if you used my link it should already be enabled. goo.gl/1JJ3Og

Mars Ultor

22:51

@Quill ?

@DanPantry wasn't for me

Dan Pantry

@MarsUltor hmm, maybe the shortlink foobarred it.

Rainbolt

@QPaysTaxes Split on regex, take the zeroth element?

Quill

@MarsUltor I can overload this probably. but not this

Optimizer

@Quill even if you can overload Object's constructor, var a = {a:4} doesn't use that

Quill

@Optimizer yeah, that's what I said. I said I can't overload the initializer declaration

Mars Ultor

22:53

@Quill Yeah, you normally can't overload literals in most languages. Also, JS doesn't even have operator overloading,

Dan Pantry

@MarsUltor (that's a good thing @ operator overloading)

Rainbolt

@QPaysTaxes This will match all text until the string FOO: ^.+?(?=FOO)

Quill

@MarsUltor Yeah, you can like fake overload it though

Mars Ultor

@DanPantry Not even for user-defined classes though

Dan Pantry

@MarsUltor Meh, I'd still prefer to not have it

Mars Ultor

22:54

@Quill But it can only return primitives, right?

Dan Pantry

though it's quite a vicious circle. the only reason I don't like it is because you rarely see it, so when you do see it it looks weird

Quill

fake overloading you can do whatever you like

there's a whole weird JS library that looks like ruby (apparently)

Rainbolt

@QPaysTaxes You don't need to be like that. You said what your problem is and apparently you have solved parts of it yourself, and assumed I would know what parts you have solved yourself.

Dan Pantry

@Quill I believe that's called CoffeeScript, and everyone hates it. /s (seriously, though, it was voted one of the worst languages in 2016 and 2015)

Easterly Irk

BAI

Quill

22:55

Def.js

quartata

I wonder what would happen if you combined Babel and TypeScript into one big ball of awesome

Maybe JS would actually be usable

Dan Pantry

@quartata It is decidedly less awesome than you think.

quartata

@DanPantry Oh, do they not play nice with each other?

Dan Pantry

  { loaders: ['ng-annotate', 'babel', 'ts'], test: /\.ts|js$/, include: /app/ },

^ been there. done that.

no. typescript is upper crap for interop with anything, including other javascript

Quill

FlowType is the way to go (if you want static typing)

Downgoat

22:57

@DanPantry +1 for babel :D

Dan Pantry

i mean typescript won't even let you interop with half of the npm ecosystem without choosing from one of three import syntaxes

all of which have different semantics

so if you try and use anything from npm you will probably end up with something like this (actual code from our app)

Quill

@Downgoat you just missed a javascript ~~bash~~ magic fest

Dan Pantry

import $ = require('jquery')
import lodash from 'lodash'
const classNames = require<any>('./classNames.scss')

Downgoat

@Quill ;_;

Cᴏɴᴏʀ O'Bʀɪᴇɴ

-1 for jquery

Is this not noise?

Transcript for