feature-request-request for Proton anyone?

const or final in Proton?

@HyperNeutrino hmmmmmmmmmmmmmmm

command chains :P

@HyperNeutrino const imo

or short builtin names at least

final sounds stupid

ok

final is also Java

@ASCII-only what is command chain

yes and java code often sounds stupid :P

:P

"workers.contains"

liek seriously it should be "workers.contain"

@HyperNeutrino chr . ord . x instead of chr(ord(x))

or just "in workers"

i think that's what command chains are

isn't that function composition

@HyperNeutrino ... what.

@totallyhuman ^^

oh

nvm me then

that's also a cool feature though

mrhaki.blogspot.com.au/2011/05/…

besides, it should be x | ord | chr

@totallyhuman well you can't exactly do that

@HyperNeutrino why not x :> ord :> chr

@ASCII-only why not?

@ASCII-only maps

@totallyhuman because . is property access?

it's used already

^^

@ASCII-only oh that was a placeholder

@HyperNeutrino wait :> is used already?

yup

dicts look like [key :> value, key :> value]

@totallyhuman but that's what every language uses for it

@HyperNeutrino ... why

lol

it looks weird ik but ¯\_(ツ)_/¯

why not do it the sane way

too bad there are so few brackets

@ASCII-only because {} is also used

@HyperNeutrino ... so?

I'm not sure it would be easy to double the meaning of {} for both codeblocks and maps. I'm sure it could be done but I'm too lazy :P

do you have sets then?

just use set(...)

@HyperNeutrino ... Of course it's easy

depends on if the parser's set up correctly, which I doubt mine is

what do you use for your parser?

@HyperNeutrino If even Charcoal can do it why can't you

@Dennis do you mind helping me out on a math question I had?

something similar to a regex engine

Also oops forgot to add colons in parser fixed now

@HyperNeutrino you what.

@ASCII-only I mean if that were the sole use of it it would be easy

@ASCII-only you're not the first to disapprove

@HyperNeutrino ... Charcoal has codeblocks too

o

I'll have to try that then

sometime soon. I have implemented dicts already so I'll do it later today :P

I want to add const first

Well really it depends on how your parser works, if you can't do that you really should rewrite it :P

@Dennis wait nvm I'll just ask it on math.se I forgot that existed

@ASCII-only you know I really shouldn't, but I kind of want to rewrite Proton to fix some major issues with it... but I mean, Proton was meant to be a rewrite of Positron and I ended up making a new language, so...

@HyperNeutrino Why shouldn't you

@Maltysen >_> TNB exists too

@Maltysen As you wish. Assuming I can, I don't mind helping you.

@ASCII-only I've made too many programming languages already and they're all bad ;_;

@Dennis oh ok. I'll try u first i guess

then again, the more I write, the less bad they get lol

@HyperNeutrino :| I only have one

What would you expect a library with the namespace inb4 to be?

I made (Practical) Positron, Proton, (Recreational) Ceres, Paintbrush, Anyfix, PostL

PostL is impossible to use for most things

@PatrickRoberts Inject methods before events are called :P (before-event hooks)

Anyfix is really long and is like Jelly but without as many commands and with way less golfability

Paintbrush is an ASCII-art language that is about twice Charcoal's length consistently

@Dennis So I was wondering how the jacobian plays into calculating the size of area elements embedded in >2d. The motivating example was surface integrals where you do |dr/du x dr/dv| which if it was a 2 var->2var mapping would be the jacobian, but since its u,v->x,y,z the jacoiban ins't square which means you can't calculate the determinat

@HyperNeutrino O_o Positron is practical too

@ASCII-only it was meant to be :P

@HyperNeutrino :| Apparently Charcoal is nowhere near as golfy as it can be

rip

@Maltysen No clue, sorry. I'm afraid I'll have to defer you to one of SE's math sites.

@Dennis the cross product works for 3d, but how do general tools like the jacobian determinant work if it isn't square. The wikipedia page for en.wikipedia.org/wiki/Volume_element even says "the volume element changes by the Jacobian of the coordinate change"

@Dennis oh ok

Oh also @anyone what would you call list = list.map(element=>element<operator><constant>) and list = list.map(element=><constant><operator>element)

@Dennis Thanks anyways

@ASCII-only prefix operator functions and postfix operator functions?

Well shoot

@ASCII-only element-wise operation

I was making a challenge and I found out it was undecideable

The internal type of something like 2+ or /2 in Proton is prebinopfunc (prefix binary operator function) and postbinopfunc

@Maltysen Kinda related but no answer

@ASCII-only Call it an "atop" :P

@HyperNeutrino if they were a function what would you call them

@ASCII-only no, this is exactly it, but I can't understand anything hes saying after "But of course I can't do this since the Jacobian is not square."

what the heck is a pull-back

@ASCII-only operator function? or something like that

not sure

Probably doesn't help either

@ASCII-only in MATLAB, the syntax for an element-wise operation on a higher-level datastructure like a vector, matrix or tensor is structure .<operator> <constant> and <constant> .<operator> structure respectively

@Maltysen Oh this may help?

@ASCII-only oooooh that's it thanks so much

@Challenger5 BTW for Charcoal addition should be vectorized too right?

Probably.

When you have your own codepage (and therefore many bytes/builtins at your disposal), then overloading takes a backseat to generality.

Proton now has const :D

uh oh BORK ALERT 🚨 BORK ALERT 🚨 BORK ALERT

um should const a, b = 1, 2 set both a and b to const?

Yes

huh this may be slightly tricky

should a = 0; const a; also do the same or should that not be allowed?

@HyperNeutrino Yes otherwise confusion ensues

k

@HyperNeutrino Nooooo

@ASCII-only ok good :3

@HyperNeutrino Don't allow declarations after definitions

k

I won't allow const a; a = 2 either then

Also function arguments should be const by default unless explicitly specified

nah I'd say not

@HyperNeutrino that should error

o ok

I have to figure out how to make a better triangling algorithm

actually I think I'll just adjust the current one

@HyperNeutrino yes definetly

@HyperNeutrino also I please make const actually consts and not just immutable references

A syntax error is a wasted opportunity.

@ASCII-only btw can you change a = 1 to AssignmentExpression and also ensure only lvalues are on left-hand side

or are lvalues all PropertyExpressions

wait

proton has unless?

no wonder it reminded me of some other language

ruby

@ASCII-only btw should we throw warnings for errors (that won't actually break though)?

Yep

example is through branch checking we can analyize the domain and if there is a 100% value fallthrough we know that it won't meet the condition for the error in runtime but it could still be error

@ATaco assignments in expressions? c'mon

Assignments should always be allowed in expressions.

^ yes

bleaugh no

But then you can do x = y = 3

what practical language would be crazy enough to all— oh right python

^^

also then say bye to for(;;) loops

it's good for golfing but horribly confusing in actual code

that is (very) false

trying to think of way to quantify confusion

while (x=get(i++)){}

is fine

got it

@totallyhuman probably not that confusing if you know the language

@Downgoat what is an "actual const"

Realisticly, Programming is confusing.

I mean [0 for i in z] probably looks gibberish to someone who doesn't know python

@HyperNeutrino meaning you cannot mutate any field/value e.g. const foo = ...; foo.bar = 123 would be invalid, exception is:

o huh I'm not sure I'd be able to do that

let a1 = A()
let b = B(ref: a1)
a1.value = ...

Does proton have Macros?

I mean obviously I can but not necessarily without modifying the underlying identifier structure

@ATaco no

@Downgoat python does not allow assignments in expressions, only chained assignments, which aren't actually confusing

:(

@DestructibleLemon no one said anything about chained assignments being confusing? (not sure what you're bringing up)

I don't see how it is crazy at all

no python doesn't do allow assignment in expressoin

well you communicate unclearly

so there

Lua doesn't allow assignment in expressions either, it makes me sad.

@DestructibleLemon thanks for the tip

@ATaco why is this needed

x = y = 3

again, that is different

@DestructibleLemon for(let i = 0; i < array.length; i++) { ... }

while x = get(i = i + 1) do end

@ATaco this is why python is good

while (line = readline())

chained assignment and assignments in expressions are two totally different beasts

@Downgoat oh so like a for i in range(len(array)):?

@totallyhuman no true pythonmen eh?

Not necessarily, chained assignments can be implemented via assignments in expressions.

@DestructibleLemon I'm not asking on how to write a loop in python

@Downgoat then why is it needed

It's never needed.

@Downgoat what?

@totallyhuman rationalwiki.org/wiki/No_True_Scotsman

Example:

If we only cared about what was needed we'd be writing assembly with punchcards.

> No good python program would need expression in statement

I think you just like the ability to link to the no true scotsmen thing

Totally human was saying chained assignments and general assignments in expressions are different

@DestructibleLemon the ability to copy paste is fascinating

That fallacy does not apply here

Assignments in Expressions is just a nice thing to have, and there is, most importantly, no reason not to have it.

this is a textbook example did you read the page

imo, the reason not to have it is confusing code

If you think it's harder to read, then don't write it.

right, but what if i have to read somebody else's code

Readable code doesn't exist, only readable comments.

^^^ if you are just learning how to read loops with assignments inside them it might be difficult to read such a program but c'mon

do you know how many times I have accidentally typed one equals sign instead of two?

no, but I have never had that mistake ¯\_(ツ)_/¯

AutoHotkey uses a single equals to match equality.

@ATaco err, what language are you defending?

I mean, not that many, but I am easily able to get it corrected because my IDE notices

I'm not defending it.

'cause most languages don't use a single equals sign

it's the worst idea, but that's because AutoHotkey wasn't designed to be a language.

s = 3
if(s = 3){
	MsgBox Why is this correct.
}

omy

I feel like an expression should just be an expression

ouch

By the way, s set to a string in that case.

not an assigning things as well

@DestructibleLemon ... And assignment is an expression

To set it to a number, one must do s := 3

Pls stop discriminating :P

@ASCII-only it's a dumb expression

I think this is the worst thing we've argued about.

@DestructibleLemon but luckily, either way, python'll throw an error in your face