I tried a different browser just for giggles.

I’m pretty sure I’m expecting the cors headers on the request to fetch http://localhost:3449/index.html

I’ll try wrapping the figwheel server and adding it.

seems like I’m doing something more fundamental incorrectly, though.

or my understanding is fundamentally incorrect

ah. that was it.

@skelter what was it?

My understanding of the focus of cors was upside-down.

I had the concern about injection of references to off-site js stuck in my head and so was expecting the cors headers from figwheel for index.html. That’s not what I should expect. I had to reread cors fundamentals, tweak my api server with cors headers.

@dnolen great thinking on the infer externs addition, very helpful!

any good reference for JS interop? e.g. a side-by-side code of implementing stuff with JS and CLJS

@ejelome http://cljs.info/cheatsheet/

thanks @pesterhazy, I meant something like this: http://norvig.com/python-lisp.html

oh found one, might be useful: http://himera.herokuapp.com/synonym.html

@ejelome many of the things listed there are out of date

but will work just fine still 😉

please someone make a 2.0 😄

is the cljs linkedin group abandoned? my join request has been pending for quite long

Probably a stupidly easy question here. When I attempt to lein minify-assets or as part of an uberjar task, I get the following error: "java.lang.IllegalArgumentException: No implementation of method: :as-file of protocol: #'http://clojure.java.io/Coercions found for class: java.lang.Character"

Well maybe it’s because lein-asset-minifier says :assets needs to be a map now.

This is what happens when you update some dependencies and try to run a year-old project as-is 😄

somebody asked me if do they need to know JS to use CLJS, and it's quite difficult to answer 😞

Interop might be difficult

I would say no if they are using one of the react bindings, as they abstract pretty well away from js interop

There’s still some basic browser stuff that would be helpful to know

yeah, knowing JS is probably gonna be helpful since all of the runtime API documentation is given in JS.

I've taught my classmates a good amount of clojurescript with a decent amount of success after they took one course that used a lisp.

without even knowing JS?

@piotr2b I think it's a bad idea to immediately present cljs with re-frame, something like reagent is more easy to digest for those who want to test the waters

@ejelome you probably need to know JavaScript basics, but I think that’s about it.

I agree @dnolen, but it's just that some people think it's possible to literally go full cljs without any sort of js

hrm yeah I don’t agree with that

you often want to call into Google Closure Library, and you need to know basic JS

but if you’re just building toy stuff with reagent

you can probably get away w/o much JS knowledge

but not for anything serious

right, and the problem arises when serious happens

on the flip side, some people are advising es6/7/8 before diving to cljs, which I think has the same effect or no?

that claim doesn’t make any sense to me

you can’t really safely use es6/7/8 (requires source to source compilation)- and regardless what does that have to do with Clojure

might be because of react?

but don't also know if it also make sense even for diving to react

doesn’t make sense to me

React doesn’t require ES6/7/8

To me it makes sense from the perspective of introducing some functional concepts if you are introducing functional concepts to a JavaScript programmer (if you need a to b comparisons, you could find them easier with es6+)

alot of misconception really

I don’t see how ES6 helps with functional concepts

it doesn’t add anything over JS from 2001

just sugar

React also doesn’t help that much with functional concepts

Yeah, but that sugar could ease transition

most people use React in an imperative way

Like the let statements, getting used to inline functions, generators for lazy seqs

It's all real hypothetical though

yeah not convinced 🙂

JS is functional the way that Scala is functional - only if you and your team really want it to be

to me that’s value proposition of Clojure, the whole ecosystem and idioms are functional by fundamental design

I'll agree with that

does anyone have any experience with tail call optimization? Having a really hard time trying to find if there are any workable solutions out there via trampoline, macro, whatever

I have to read and parse through a very large file line by line and am concerned about stack overflow if I do it functionally using recursion

@mattsfrey any reason not to use the loop structure?

does loop/recur more or less achieve that?

I'm not familiar with the underpinnings

That's exactly what the Haskell/Purescript people say to me with Clojure 😂

@mattsfrey loop is clojure's way of doing recurrsion since the runtime doesn't natively support tail-call optimization

(call it a workaround if you will....)

loop is clojure’s way of doing TCO recursion, specifically, forcing it — it will do regular recursion just fine without loop

right

oops, mixed up the terms a bit there~

no worries, i figured @akiroz just wanted to be clear

I suppose thats the answer for now then, I had heard rumors of es6 TCO support at some point but haven't been able to dig up anything fairly modern about that development

@akiroz yes Haskell/PureScript focus on restraining mutation even more - but I remain unconvinced it matters much

people keep saying this typed functional thing is going to take off - but when people look for types they just want Java - thus TypeScript’s popularity

https://github.com/lukehoban/es6features/blob/master/README.md#tail-calls

That support is browser dependent though, so just use loop/recur and let the compiler figure it out for you

anybody used react-native-maps (https://github.com/airbnb/react-native-maps) with “re-natal" with some success ?

bRepl + eval last sexp is magical

I don't know what else to say

@qqq give Figwheel a try at some point 🙂

speaking of figwheel how much work would it take to add readline support. I’m looking for a project to contribute to.

I've tried figwheel before

brepl + eval last sexp is actaully more imprssive imho

@7h3kk1d I normally just use rlwrap

@anmonteiro That’s what I’ve been using too but it seems to act up sometimes.

I might just have something messed up locally though.

How do I escape the double quote character ” in CLJS?

”\”” seems to include the backslash when used in a string/replace call

@talexxx can you provide a concrete example?

the backslash is the way to do it normally

cljs.user=> (clojure.string/replace "ca\"t" "\"" "n")
==>"cant"

Need to prefix with a # to your 2nd arg

@talexxx can you provide exactly what you're doing? Because just doing "\"" seems to work?

(regex literal)

(string/replace ""Foo and bar and things" means" #"(")" "\"")
"\"Foo and bar and things\" means”

(clojure.string/replace "ca\"t" #"\"" “n”)

@mfikes the # isn't actually required for replace, but it works the same either way

(in this case...since I want an exact match and am not doing anything regex-y)

Actually, that might be fine?

I assume the output is escaping it so that the result is valid

@talexxx yep, if you call println on that, you should see the \" go to "

Oh, true

cljs.user=> (goog.string/unescapeEntities ""Foo and bar and things" means")
"\"Foo and bar and things\" means"

Thanks