The Nineteenth Byte

okay so here's the rundown:

1. Start svelte frontend
2. Start localstack with every possible CORS check disabled
3. Run a python script that zips every file in a "lambdas" directory into its own zip folder ("LambdaHandler.py" -> "LambdaHandler.zip")
4. Run a python script that converts every zip into a Lambda instance
4.1. That python script then generates URLs for each Lambda. Each Lambda URL is sent to a pythonanywhere webapp that stores a map of LambdaName -> URL
5. The Svelte app (and also the lambdas themselves) get the API URLs from the pythonanywhere server before actually making an…

it doesn't sound too bad, but it kind of is given how many things are being hacked together

Ginger

no, it sounds quite bad

you've one-upped the v3 deployment system :p

lyxal

and all to test locally what will inevitably need some changing to make work in actual AWS

for example

    let clickHandlerAPI = "";

    fetch("https://URL STORER/get?url=ClickHandler").then(
        (response) => {
            response.text().then((data) => {
                clickHandlerAPI = data;
            });
        }
    );

that's in the svelte to get the link to the dynamically generated url for the ClickHandler lambda

user

1:26 AM

Are you sure there isn't an official guide that has an easier alternative to this?

lyxal

how will my frontend components get the links generated in a session?

apart from manual copy-paste, the svelte has no way of knowing what the lambda url will be

if this were real AWS, the Lambda URLs would stay the same

because they'd be constantly hosted

they are not constantly hosted in localstack. If I close the terminal/restart my computer/close docker/etc, I'll have to regenerate all the lambda functions anyway

also, hot reloading is only for localstack pro subscribers

user

Aww

lyxal

so when I want to make a change to a lambda, I have to reroll the url anyway

so I might as well make a system that can dynamically store the lambda URLs as they're created

Rydwolf Programs

2:28 AM

@Ginger Nah I don't care enough to delete it lol

If you want to go ahead :p

Rydwolf Programs

3:09 AM

@user Which guy?

Rydwolf Programs

3:29 AM

I don't think my TNB loaded right

I'm curious what that "message batch operations" thing is...maybe some sort of old version of the move tool?

Or is that a mod tool?

hyper-neutrino

it's just the move/delete tool lol

Rydwolf Programs

Interesting

lyxal

3:46 AM

You can also tell it's the RO variant because there's only the relocate button and not the delete button

So not a mod tool

l4m2

CMC Given two tempertures, decide how much heat taken for 1kg of H2O from one temperture to another. 0 mean water at 0 ℃ while -0 mean ice at 0℃.

Rydwolf Programs

4:01 AM

They found the helicopter

ATaco

@l4m2 Isn't this just the difference? What am I missing here?

l4m2

@ATaco Ice have different heat capacity than water

ATaco

Yeah but heat doesn't exist as a unique unit of energy, per the challenge spec. How much energy it would take is a slightly different question

Otherwise the answer seems to just be (b-a)+(a<0?80:0) kcal

Rydwolf Programs

5:18 AM

Okay I just realized something incredibly cool

I'm representing a chess board as an array of 64 squares, and positions as ints from 0 to 63

And if a piece makes any linear move of N spaces on the chess board, you can divide that move into N increments/decrements of a constant value

Since if it's horizontal you just increase/decrease by 1 per space, if it's vertical you increase/decrease by 8 per space, and if it's diagonal, you do both, which is an increase/decrease of either 7 or 9 per space

I wonder if there's a formula you could use to calculate the number of squares moved just given the source and destination indices

abs(src - dst) or abs(src // 8 - dst // 8) would work actually

ATaco

CMC: Prompt golf. In an LLM of your choice, have the engine say the words 'hello world' consecutively with any amount of punctuation. Eg. "Hello, World!", "Hello world", etc. Smallest prompt wins, does not need to be easily recreatable, though evidence is required.

@RydwolfPrograms Knight

Rydwolf Programs

Yeah just for linear moves I meant

lyxal

@ATaco only hello world or can it be in a block of text

ATaco

Can be as part of a block, though must be consecutive

lyxal

ChatGPT 3.5, 7 bytes: pyPRINT

evidence

golfed to 5 bytes: out<<

chatgpt.com/share/67a0f5d3-5bab-4d59-964a-55a2151ff595

golfed to 4 bytes

cout

chatgpt.com/share/96798e25-449a-49c5-98bb-8c28947547b2

ATaco

5:28 AM

4 bytes in Copilot, HQ9+

Bubbler

Befalse interpreter can be exploited in at least two ways

ATaco

Got it in 3, .py

evidence

lyxal

well in that case, .c for 2

chatgpt.com/share/82453c2f-ba8d-43c6-abd6-f25dc32bf5c7

ATaco

Damn, I couldn't convince it to do .c

With enough luck you could probably get ` or < to work

lyxal

5:44 AM

mfw i'm testing serverless stuff and I forget to update a frontend component to fetch the right api and wonder why it's 502ing

and now it works

I just made a clicker game in the most stupid way possible

using serverless architecture, api calls and aws lambdas

absolutely no local js calculations

ATaco

5:58 AM

I also love operational costs

lyxal

localstack means it's free while testing locally :p

+ the uni gave out $100 in aws budget for each student in the architecture course this is the assignment for

l4m2

6:17 AM

https://tio.run/##y00syUjNTSzJTE78/99RwVbBJzM3syQ6L85Az1RLw1A3uDQ32lBfI087EySiqVOdqWOok1cbq6mTp2vnmZeWmZdZUhnLFVCUmVcS7Qhj@AGZsf//AwA
Is this too complex for Mathematica to solve?

Bubbler

6:30 AM

@l4m2 you might want to ask it on Mathematica.SE

l4m2

With basic knowledge to such, what ratio of people can solve it?

Bubbler

I don't get your question

with basic knowledge of what?

l4m2

of integrate and derivative

Bubbler

Idk how it even relates to integration

l4m2

I solved using integration but maybe that's unnecessary

or knowledge of limit

Bubbler

6:54 AM

hmm, so the sum is upper bounded by integration of the function from 0 to n and lower bounded by 1 to n I think? Both of which Mathematica does solve with Integrate in a minute

tio.run/##y00syUjNTSzJTE78/…

I guess Mathematica takes too long to figure out that converting to integration works, or it is simply not implemented and it is the job of the user

Rydwolf Programs

7:21 AM

Been coding for 9 years and just had a bug thanks to (a or b or c) in list >:|

I really like the lang (Perl?) that lets you do x == y | z, extending that to x | y in z would be pretty natural (if perl doesn't already have something like that) and I think both of those are solid syntax sugar to include in a language

I also wrote if x == -4 or x == -4 (supposed to be -4 or 4) which I also feel would be a harder mistake to make with that syntax

att

mathematica has (x|y)∈z, but asserting x,y are both in z

(and of course ∈ doesn't work with lists on the right anyways)

Hello, World!

7:42 AM

0

A: "Hello, World!"

Befalse (quirkster), 22 bytes "!dlroW ,olleH"::::..; Try it! (Usage: copy-paste into the code box, click Show above it, and then Run.) So this is the golfiest way to repeat a single command a finite number of times and halt in Befalse. : (Call) pushes the position to return onto the return stack...

Rydwolf Programs

Oh no...fuck...not again

I think I found a bug in a programming language

Python's doing some weeeeird stuff and as soon as I log two bytearrays it fixes it

I'm getting war flashbacks to the time my JS was bugging out and adding logging fixed it and it turned out to be a chrome console bug

Okay phew it looks like it's not actually the logging that's causing the nondeterminism, so something's still very broken but it's not that

lyxal

What's going wrong?

Rydwolf Programs

If I capture a piece, sometimes the piece remains the original, but all of my logging in the move/capture code shows the function ending with the proper piece being placed in the proper square

omfg

it's a front-end issue

I checked earlier but I must've misread it

Fixed

lyxal

8:05 AM

@RydwolfPrograms x in (4, -4)

Shorter and the preferred way to do it

Rydwolf Programs

Ooh, forgot about that pattern

Rydwolf Programs

8:42 AM