02:35 <jfhbrook> jesus I haven't even written a line of typescript and I'm already seeing this glaring problem

02:36 <jfhbrook> there's no culture of wrapping javascript libraries to have stricter typing, at all

02:37 <jfhbrook> so it's like, oh I'm getting an object from minimist that *I* know has a structure, but the typing system actually knows very little about it? string keys and a union of value types? great!

02:37 <jfhbrook> is there an options parsing library that really leverages this new type system? no! and people don't even really see what the problem is!

02:38 <ec> )=

02:38 <jfhbrook> We Don't Make Any Money If You Don't Click The Fucking Link

02:38 <ec> see? I'm not an illogical hater, I swear! I like TypeScript! I just like it for _one specific task_, which other, better tools chose to be incapable of fulfilling.

02:39 <ec> problem is, the only other viable option is BuckleScript (Sorry, Flow, you lost the war, apparently. And Fable's too young. And ghc.js, jesus christ, what a nightmare.) — which also has all those aforementioned issues

02:40 <jfhbrook> flow is apparently really bad

02:40 <jfhbrook> super buggy

02:40 <ec> this is precisely why I'm putting so much time into BuckleScript 'n OCaml crap. I really, genuinely, honestly care about the JavaScript ecosystem; and I really, really think we're all screwed if BuckleScript/Reason/whatever dies.

02:40 <jfhbrook> yeah, it's looking like either we use typescript and make an active effort to make a walled garden, or we use purescript or bucklescript and are forced to write a bunch of bespoke code

02:41 <ec> fwiw, I forgot to mention this, but Elm can be really great, if you're planning the "walled garden" approach.

02:42 <ec> Evan makes, uh, large, sweeping changes that piss people off on the regular, tho, and yeah, needs to be — again — as agnostic to external libraries as possible; I hear interop is a nightmare.

02:43 <ec> got to talk to more than one group that's using Reason in production *not even because they like Reason* — but because they bought into Elm, and got fed up with that stuff, and ended up deciding BuckleScript provides better/looser/more-escape-hatch-friendly interop.

02:43 <jfhbrook> yeah I flipped through the elm docs and idk I got a bad feeling

02:43 <jfhbrook> bucklescript and reason make sense to me, philosophically

02:43 <jfhbrook> reasonml*

02:43 <ec> ugh thing is, I've only ever heard Reason users hate on Elm, lol, so. grain of salt.

02:44 <ec> literally everybody else I've spoken to about it _loves_ it.

02:44 <ec> ¯\_(ツ)_/¯

02:44 <ec> it's all a mess and I'm tired and I really, really just want _some_ ML-derivative to succeed and survive alongside TypeScript.

02:44 <jfhbrook> well like my cofounder (potential cofounder) originally suggested purescript and was interested in reason being more approachable than something that's clearly internet haskell

02:44 <jfhbrook> you know what we *could* do lol

02:44 <ec> I know absolutely, literally nothing about PureScript — how closely does it track to JavaScript runtime semantics?

02:45 <jfhbrook> oh I know very little about it as well, but I *think* it's purescript : haskell :: reason : ocaml

02:45 <ec> because the lazy-evaluation stuff is an absolute nightmare w/ ghc.js. diametric opposite of BuckleScript, re: output being semantically clearly linked to the input, lol.

02:45 <jfhbrook> oh, I see what you mean - it sounded like it was meant for writing practical software in the same way bucklescript is

02:45 <jfhbrook> but I haven't done the reading

02:45 <jfhbrook> I like *just* remembered what it was called last night before bed

02:45 <jfhbrook> and then focused on researching typescript tonight

02:45 <ec> my understanding (hope?) was that PureScript was a _new_ language, only similar to Haskell in syntax and type-level concepts, but with a more Web/JavaScript-friendly runtime evaluation model

02:45 <jfhbrook> after a **hilarious** day at work

02:46 <ec> so what's this thing you're gonna build, anyway?

02:46 <jfhbrook> that sounds likely ec, it didn't look like it was trying to actually be haskell

02:46 <jfhbrook> ooooh

02:46 <jfhbrook> so

02:46 <jfhbrook> we're still trying to figure out how to pitch it and exactly what it looks like, but here's how I'd explain it to an industry person

02:47 <jfhbrook> so we want to build the capabilities of a content studio and/or an ad agency - we have professionals on staff that can write, record, and anything else content-related you would need; but we target the long tail, small time stuff, using techniques similar to gig sites like taskrabbit or fiverr

02:48 <jfhbrook> the mvp is more claire's original pitch: an app where you can easily work with a professional writer to craft small pieces of text, like a difficult email or a brand tweet you want to go viral

02:48 <jfhbrook> so we make it easy to buy small amounts of text written by people with bylines in things you've heard of