The htmx edition.

>Free beginner resources to get started with HTML, CSS and JS
https://developer.mozilla.org/en-US/docs/Learn - MDN is your friend for web dev fundamentals
https://web.dev/learn/ - Guides by Google, you can also learn concepts like Accessibility, Responsive Design etc
https://eloquentjavascript.net/Eloquent_JavaScript.pdf - A modern introduction to JavaScript
https://javascript.info/ - Quite a good JS tutorial
https://flukeout.github.io/ - Learn CSS selectors in no time
https://flexboxfroggy.com/ and https://cssgridgarden.com/ - Learn flex and grid in CSS

>Resources for backend languages
https://nodejs.org/en/learn/getting-started/introduction-to-nodejs - An intro to Node.js
https://www.phptutorial.net - A PHP tutorial
https://dev.java/learn/ - A Java tutorial
https://rentry.org/htbby - Links for Python and Go
https://quii.gitbook.io/learn-go-with-tests - Learn Go with Tests

>Resources for miscellaneous areas
https://github.com/bradtraversy/design-resources-for-developers - List of design resources
https://www.digitalocean.com/community/tutorials - Usually the best guides for everything server related

>Need help? Create an example and post the link
https://jsfiddle.net - if you need help with HTML/CSS/JS
https://3v4l.org - if you need help with PHP/HackLang
https://codesandbox.io - if you need help with React/Angular/Vue

/wdg/ may or may not welcome app development discussion. You can post and see what the response is.
Some app technologies of course have overlap with web dev, like React Native, Electron, and Flutter.

We have our own website: https://wdg-one.github.io

Submit your project progress updates using this format in your posts, the scraper will pick it up:

:: my-project-title ::
dev:: anon
tools:: PHP, MySQL, etc.
link:: https://my.website.com
repo:: https://github.com/user/repo
progress:: Lorem ipsum dolor sit amet

Previous: >>105787235

Give me open source 2ch-style text boards like Kareha but less complicated.

Text only, I'm not interested in imageboards

Guys I really like MariaDB

It's legit taken me over a month to refactor part of my codebase to switch from runtime acquisition of certain data on a per worker basis in a distributed queue to making the workers stateless. An absolute fucking nightmare made worse by the fact that I have a python job queue and a ts job queue and I had absolutely no other choice but to do it like that.

>>105872088 (OP)
TS >>> JS >>> HTMX

>>105872283
Good lad. Fuck Oracle.

Haven't seriously checked front end frameworks in years, does React still have the richest ecosystem or did others catch up?

>>105873894
react is deprecated

>>105873894
Yeah React probably does but Angular and Vue are pretty widely used and some people love Svelte and other frameworks. I just use React because it just works.

>>105872088 (OP)
any K&R tier books on TS/JS? i use them a lot at work but routinely step on rakes whenever i do because they do things different than C/C++/C#.

>>105874619
eloquent js

anyone else here learning with chatgpt? I'm asking it for js challenges and shit

>Be me, try React
>Create controlled input element on jsx
>type="number"
>put numbers in input (3.1415, 69420)
>Do mafmagics with number
>Entire component crashed because state tied to input is now a string :3

I am filtered already

>>105876191
I use it to check for syntax errors and explaining specific topics but avoid relying on it for writing any code

>>105877514
next stop: typescript

>>105872250
I wonder if Claude Opus 4 could one-shot this

https://github.com/vuejs/core/releases/tag/v3.6.0-alpha.1
>vue vapor mode in alpha
sveltetrannies your time is almost up

page 9

>>105881019
Based thread bumper

>>105881197
you too

>>105872088 (OP)
First time running a domain here. I never expected there to be so many bots.. Let alone that most of them are European and acting in bad faith! Like looking for .env and wp-admin pages
A few thousand requests from france and the nederlands in a week... just fuck off
Although I think it's pointless to do so, but should I ban clients that act maliciously?
Is there another way to deal with this?

>>105872088 (OP)
I spent some time enjoying/learning Svelte, but finding a plain Svelte template proved difficult. I keep getting npm and Vite errors.
SvelteKit seems like overkill for just a frontend. I have to disable ssr everywhere, enable pre-rendering, and be careful not to use server-side features or functions that might run on the backend (like page.ts/.server.ts files). It's a lot of extra work just to avoid potential issues e.g. online guides says put `export const ssr = false` in this file and then I get a warning telling me to put it elsewhere..
Is there no framework that just goes name.format => name.html/css/js + vendors/fonts?
I just need to some content on a template, but most ssg are either too complex or too limiting.

>>105881765
have you tried astro? you can bring whatever framework, even multiples of them. their .astro files are a bit weird, kinda like JSX + Frontmatter but you get used to it.

I've written back-ends in a few languages and now I'm thinking I might try a back-end in C++. Is this a dumb idea or a great idea?

>>105882260
yeah do it

>>105872088 (OP)
>>Free beginner resources to get started with HTML, CSS and JS
>>Resources for backend languages
>>Resources for miscellaneous areas
Does it even make sense to read a book instead of learning by vibecoding nowadays?

>>105882260
Webdev work is mostly IO bound. What difference does it make if your node/express takes 15ms to handle a request compared to C++'s 10ns of manualy writing HTTP/1.1 requests? Even a single, typical, plain hit to the DB or even checking if a file exists dwarfs the CPU time needed by your backend. If the resources you need are not cached or physically close, nothing you do really matters.

>>105882316
It hardly made sense before, why would it now? Learn the basics from a cheatsheet or quick start guide then start building and fixing issues as they come. Ask AI to help find the right words/terminology for a problem/solution. Learn to make use of prebuilt & trusted solutions to solve your problems. Don't make everything blocking your way "your problem", install a package and lock in on your real focus.
Life taught me that thinking everything will flow your way once you finish a book or course is extremely childish, especially for fast moving fields.

pls help a clueless JSlet, webfrens. I wrote a bookmarklet for Firefox which is just a JS switch statement that sets window.location.href to a different URL depending on the output. It works fine for most of the stuff on my domain except for the qBittorrent webUI. It'll correctly redirect to mydomain.com/qbit/ and load the UI to the login screen and just loops that page, no matter whether I enter the credentials, refresh it or even hard refresh. The only way to fix it is if I hit ctrl-L to select the URL and then enter. I've no idea why. The only thing I noticed different is the "initiator" tab in the network devtools changes from the bookmarklet code to "document" when I ctrl-L+Enter. Anyone have any idea what I'm doing wrong?

>>105882295
I might just dabble just because I'm curious about C++

>>105882340
Yeah fair enough. I'm just curious about C++. For the last few years I've only used GC languages. I feel like I should learn new stuff.

Vibe coded a tampermonkey script for clicking next episode on watch-cartoons-online, so I don't have to click the next vid on bluey or peppa pig for my 2.5 years old. Unfortunately, the site owner has prevented autoplay. Clicking the button does nothing and executing play() on the player serves a dummy video. I assume the server is checking for event.isTrusted

// ==UserScript==
// @name Wco Main
// @namespace http://tampermonkey.net/
// @version 0.1
// @description
// @author
// @match *://www.wcostream.tv/*
// @match *://wcostream.tv/*
// @grant none
// ==/UserScript==

(function() {
'use strict';
window.addEventListener('message', function(event) {
if (event.origin === 'https://embed.wcostream.com' || event.origin === 'http://embed.wcostream.com') {
if (event.data === 'videoEnded') {
document.querySelector('a[rel="next"]')?.click();
}
}
});
})();

// ==UserScript==
// @name Wco Iframe
// @namespace http://tampermonkey.net/
// @version 0.1
// @description
// @author
// @match *://embed.wcostream.com/*
// @grant none
// ==/UserScript==

(function() {
'use strict';
function setupVideoEndedListener() {
const video = document.querySelector('video');
video?.addEventListener('ended', function() {
window.parent.postMessage('videoEnded', '*');
});
}
setupVideoEndedListener();
})();

Vibe coded a tampermonkey script for clicking next episode on watch-cartoons-online, so I don't have to click the next vid on bluey or peppa pig for my 2.5 years old. Unfortunately, the site owner has prevented autoplay. Clicking the button does nothing and executing play() on the player serves a dummy video. I assume the server is checking for event.isTrusted

// ==UserScript==
// @name Wco Main
// @namespace http://tampermonkey.net/
// @version 0.1
// @description
// @author
// @match *://www.wcostream.tv/*
// @match *://wcostream.tv/*
// @grant none
// ==/UserScript==

(function() {
'use strict';
window.addEventListener('message', function(event) {
if (event.origin === 'https://embed.wcostream.com' || event.origin === 'http://embed.wcostream.com') {
if (event.data === 'videoEnded') {
document.querySelector('a[rel="next"]')?.click();
}
}
});
})();

// ==UserScript==
// @name Wco Iframe
// @namespace http://tampermonkey.net/
// @version 0.1
// @description
// @author
// @match *://embed.wcostream.com/*
// @grant none
// ==/UserScript==

(function() {
'use strict';
function setupVideoEndedListener() {
const video = document.querySelector('video');
video?.addEventListener('ended', function() {
window.parent.postMessage('videoEnded', '*');
});
}
setupVideoEndedListener();
})();

>computer stops working
>"same shit as always"
>remove ram
>put 2 sticks out of the 4 back
>boot
>"justwekrs"
>turn off
>put the rest of the sticks
>works
this has happened to me like 6 or 7 times in the last 8 years, always same exact stuff, except normally it'd undo my ssd raid, and this time it didn't new ram sticks

>>105883496
Have you tried turning it on and off again, unironically? Booting twice seems like requirement when changing RAM for me.
Otherwise, it's either RAM stick or slot issue.
I typically only install RAM that comes in kits and don't mix different ram sticks. I think the similar issue I had was dual/single channel issues where individual RAM sticks (of the same model!) can be either dual or single.

Page 9

What do you guys think about fixi.js? It's the light version of HTMX.
https://github.com/bigskysoftware/fixi

>>105883212
just css the site so your kid can only see the relevant buttons. bonus points for increasing difficulty in button finding.

>>105885678
Why shouldn't I just use JS/TS. I already know JS/TS.

>>105885678
javascript frameworks has entered the lobotomy era

>be GPT chud noob
>find out how models, controllers and routes actually work after overhauling them thrice
lol

What do you guys think of this vanilla js framework?

It's built to just werk with the browser without any build steps, but i've been using it with vite and it's pretty nice

https://www.npmjs.com/package/basedom
>no dependencies
>signals and effects
>tinybase included, but tinier
>offline applications are first class

>>105883212
I make next button and theme edit and subtitle customizer scripts all the time for manga and anime sites.

>>105888189
Peoples will be turned off by flutter style of constructing UI, it make ugly code.
JS just cant pull it off as good as swift and kotlin
JS needs lagged template literal like lit-html

>>105888651
>lit-html
It's compatible with lit-html afaik
The dev has considered this issue and has some ideas on how to address this.

I've done plenty of web dev but it was a while ago and I've never done hosted stuff. I was going to make a site for fun, is there a tl;dr on hosting and what kinds of dev tools people use nowadays? I'd ask an AI but I have no idea if the answer will be decent or not.

>>105887673
Lmao even. Just spent two hours mopping up, didn't get any new progress done.
At least I'm catching on by the fringe. I think I can stop editing soon and start implementing base logic that any webtroon should know by now.

Any webtroon worth his socks*

>>105888189
if you want a framework that works without a build step, just use vue

>>105888189
For me
>*CRACK*
It's the MERN Stack
>*Siiiiiiiip*

>>105888883
Vue has a build step.
Browser doesn't handle .vue files naturally.
Also, jsx and typescript aren't native to the browser and require a build step.

There's a lot more purpose to a vanilla JavaScript framework than no build step.
Like no dependencies, readable/editable code, just werks in the browser

>>105889077
vue can be used without a build step
https://vuejs.org/guide/extras/ways-of-using-vue#standalone-script

>>105889123
BaseDOM primarily serves these needs:
>Lightweight library for small projects.
>Readable, simple, easy to understand, open source
>Fine-grained reactivity over a virtual DOM.
>Avoid complex build setups
>No dependencies or hidden magic
>Value simplicity and minimalism over a feature-rich ecosystem.

Eventually as this project gets more feedback from actual developers who see this as a close fit to their needs it will mature and become more competitive with pre-existing frameworks.

The core philosophy of the project is minimal dependency, maximal utility, without sacrificing simplicity. Offline SPA's are first class.

>>105872088 (OP)
Have an interview tomorrow, will be cramming all day today. Wish me luck boys

>>105889123
It would be more accurate to compare BaseDOM to petite-vue or Alpine

https://github.com/vuejs/petite-vue

>>105889077
If you're websight is for hobbyist then esm.sh builds your jsx so (you) dont technically need a build step

>>105888889
>MERN
literal basedcaf stack
try DAM
>Django
>And
>htMx

>>105889312
>basedcaf
Wut

>>105889309
That's literally a build step lol
Also defeats the whole purpose of BaseDOM, which is meant to get the dev closer to vanilla js and html and avoid dependencies other than the browser itself

Vanilla JS and HTML are the future
fuck vue, fuck angular, and especially fuck react
HTMX and Fixi are the only good projects I've seen mentioned in this thread
and I think I'm switching to Fixi now because it makes HTMX seem like bloat in comparison

>>105889331
this gay troon website wont let me say the S word
its a bean
s
o
y
c
a
f

>>105889347
>-caf
https://youtube.com/shorts/mQ-9NPtsh0c?si=rh3zpyaRuFiyFyNG

>>105889377
tell me you drink starbucks gay-milk coffee drinks or my boner will die

>>105889385
Oh...like decaf...I thought you were referencing an established meme...while instead you're whiffing
Hard.

>>105889400
its not my fault this gay website censors the word for a gay bean that people milk and put in their caffeinated drinks

>>105889412
Ooohh so it's a trend
Mildly acceptable memeage
Well it's certainly not going to stop me from finishing my gigachad MERN stack
It's taken me many Vyvansepills but I've finally learned ChatGPT

>>105889418
no bro not chatgpt no no not like this
you gotta at least use co-pilot, gemini cli, deepseek, claude, and grok along with chatgpt to ensure you sift out all the bias from each one
then you take all the outputs that you've been logging and feed them to your fine-tuning pipeline so your copy of DeepSeek R1 running on your gigachad homeserver with 4 A100 GPUs can distill the learning and become familiar with your codebase
then you can hook it up to anon kode and drop gemini cli because you'll have a localized version of claude code but with better reasoning, less bias, and more closely align with your codebase, linguistic style and opinions

https://roundhere.net/links/2025/03/anon-kode/

>t. gigachad dev who drinks mocha's with basedmilk in the morning

>>105889447
I only use ONE language at a time
Not a fan of all these tranny frameworks

>>105888796
I would recommend using a VPS. You will have to learn some stuff, but it's not that hard and it's evergreen knowledge. It's much cheaper than using the various one-click deploy services like Railway or Vercel, and it's also much cheaper than AWS (not really one-click). Most people don't know this but AWS and the other big cloud platforms are very expensive.

Here is a good playlist about self-hosting on a VPS:
https://www.youtube.com/playlist?list=PLLnpHn493BHHAxTeLNUZEDLYc8uUwqGXa

There are many good VPS providers. I have used Hetzner for many years now because they are the cheapest and have good uptime. You might have to send them your id (most VPS providers will require this). They will ban you if you're doing anything shady (illegal or if you're breaking their TOS) and they are very strict about it. I have never had any issues but I see some people complaining about it. Digital Ocean is another solid choice and there are many other good choices as well. It doesn't matter that much who you go with. The beauty with knowing how to deploy your web app on a VPS is that you can easily switch to some other VPS provider or even use a server you own yourself. That is not the case if you use one of the big cloud providers.

Hetzner vs digitalozean anyone?

>>105889541
aws
>inb4
stop being filtered

for me, it's having my employer pay for claude max x20 and using claude code to do my work while i play vidya

>aws
Biggest grift in tech.

>>105889553
why would I want aws for simple hosting? isn't it more expensive?

Good interview with DHH about why they left the cloud.
https://www.youtube.com/watch?v=a30vFpSaoZg

>>105888889
I think I prefer PERN

got a question, maybe it is stupid but I am still learning.
I want to build an app in react and also want a frontpage or some sort of one page intro for it with a start button to get to the app.
My thought was to just yarn create vite frontend, chose react. in this created react project I would do something like /app and /landingpage inside the /src folder. is this okay or would I do something stupid here?

>>105889479
>It's much cheaper than using the various one-click deploy services like Railway or Vercel
What do you mean? Vercel can host a small website for free. Backend is also free on a ton of providers if you take the cloudpill and make it serverless.

>>105889461
it is one framework
transformers are all you need

>>105880031
Wow, they ported alien-signals AND vapour mode.
Nuxt chads are also eating good, Nuxt 4 is about to release (for real this time).

Anyone know how I'm supposed to design icons for my web app gps? I want to create a boat, plane, car and stickman like the way it looks on google maps. Do I actually have to design these myself? My design abilities are atrocious and it would take way too long, I don't have time to do this shit.

>>105882741
maybe try this instead of location.href https://developer.mozilla.org/en-US/docs/Web/API/Location/replace

>>105890644
We use those where I work https://tabler.io/icons and if there is something which doesn't exist we make them ourselves in figma

>>105890644
ewither find an iconset, diy, or pay someone to do it

*devtoddles behind you*

Page 8

>>105872088 (OP)
so... what's the deal with htmx?
is it enough to make a .net mvc project viable?

I know some C# so I thought I would take a look at how people make websites with this and read about Blazor

man, what the fuck is this shit? It feels like a gigantic clusterfuck

>>105892239
lol i was forced to use blazor (ssr) for a new feature in my company, in fact i still use it
the idea was to make c# backend oriented devs able to quickly build simple interfaces
and it works fine, until you need to do something more complex than a form
also if there's even a bit of latency, you're gonna have a nasty experience with signalr

>>105892012
Give it a shot and let us know how you get on

I need to create a web app for any sort of business of my choosing and I need to log and track all my code using github and draw use case diagrams. My app can't be something basic but at the same time I need something that won't be too difficult for me. I need something easy enough to pass but not to simple to where its not good enough for a pass. I have a month to complete this project.

>>105890311
>trans

>>105881583
Use fail2ban, it scans your server logs using regex matching and temporarily bans bots accordingly

I am a backend retard who enjoys HTMX.

Is there still a way to make some money on the side on platforms like Upwork or fiverr? Or is it over?

>>105894360
>backend retard
for me back ends are chads

>>105895391
My brain works like I prefer to have a few simple tools that I can arrange in countless ways to solve a problem, rather than have a bunch of specialized tools that do one thing well. In other words I prefer building things rather than learning things. Probably worse in the long run for productivity and keeping up with industry best practices, but whatever.

>>105895498
yes, you either become proficient at DIY, or you fall victim to "industry standards", aka, fw of the week

>>105895498
Just use AI to replicate industry best practices for your use-case.

Name one reason Mongo still exists now that everyone codes with ORMs anyways

>>105890267
>make you're backend in honojs
>deploy it to serverless
>when it comes to pay the serverless jew move it to ec2
Unironically what's the catch?

>>105898305
I don't understand the hono meme.
>Just use a thing made from bun's HTTP server which is just uWebsockets.js even though uws is faster on node
>no you don't understand hono is made to be like express even though there's node packages made from uws that already do that like hyper-express and ultimate-express without getting memed on by bun

You guys watched a fireship video and thought the javascript faster than rust meme was real lmao

>>105898305
>>105898393
Ultimate Express is literally a 1 for 1 drop in of Express, the API is exactly the same. You can just change out the import.

According to the benchmark from the creator of Elyisa, a bun framework, uws and ultimate-express practically smokes everything else including his own framework and bun itself.

The kicker is the test itself is kind of stupid because the Hyper-Express creator did his own benchmark over network and showed that HE is only barely below uws.js in the real world.

The Elysian creator did his test over localhost which is absolutely retarded.

>>105898393
Does hyperexpress run on cloudflare workers? Hono is made to not depend on node.js
And also to keep the js footprint super small for cold start issue

>>105898644
>serverless
Do not talk to me

>>105898865
>paying when it could be free
Good goy

jonas schmedtman course is good for beginners? i want to learn js, react. thank you

>>105898928
>course
SAAR

completely vibe coded a batch file renamer gui with electron after uploading a jpg of the windows equivalent

the future is here boys

>sneed the database already goyim!
I'm trying.
Fuck.

Anyone know what type of web app I can build in 2 weeks that is intermediate level. I need to showcase something for college in 2 weeks but it can't be some simple pile of shit. How difficult is it to learn and utilize apis in 2 weeks?

>>105900745
>How difficult is it to learn and utilize apis in 2 weeks?
APIs are basically functions over the network.

They have a function name (URL path), function parameters (GET / POST body), and can return something back to the caller (JSON)

As long as you understand that, you shouldn't have much issue implementing or using one. Maybe the other stuff on the side like databases perhaps.

with regards to a web app that you could build in 2 weeks, you could try something like a chatroom plus a synchronised youtube video playlist using socket.io, pic rel is what I did literally 10 years ago

For shit and giggles tried Haxe with its react-haxe lib. React 18 is not (yet) supported, the createRoot stuff, but feels good being able to just transpile existing Haxe libraries, and use it for whatever the fuck you want. Also, wrapping native Javascript libraries is piss easy, and you get the added benefit of type safety.

>>105901251
Pretty solid project for 10 years ago mate.

>>105901251
I need something for hypothetical commerical use not something a 10 year old would write. Reminder that I have wrote some static webpages in the past but that was some time ago so nothing too difficult now.

>>105890267
It quickly gets very expensive when you get some traffic and you're locked into the cloud/"serverless" when that happens.

>>105898305
>ec2
Still very overpriced compared to running your own servers.

>>105903089
Read his post again lad

>>105903089
APIs have real commercial use. Entire business are built around API endpoints.

PSA: applets have been removed, please migrate
https://github.com/openjdk/jdk/pull/25698

>>105907198
Good, fuck that legacy shite

>>105907938
runescape used them for example, but browsers have long removed support. the idea was flawed and Java mostly moved to the backend, but now that the remnants of the applet code are gone, the era feels officially over now

>>105882239
Thanks anon

>>105901251
Is that a synchtube clone I see? I got one up myself right now at https://simul.social/r/taylor

>>105883212
Maybe try doing it with MutationObserver? Does anything in the DOM change when the video ends? If yes it should be possible.

>>105883212
>>105910071
I am stupid.
You said that clicking doesn't work, not the end detection. Maybe try targeting the child or the parent elements? It sometimes works. If it doesn't the last resort would be to use AutoHotkey to do the clicking. Make the script write to clipboard when the video ends and make AHK wait for clipboard before clicking.

>>105883212
>Unfortunately, the site owner has prevented autoplay

It's not usually that, it's usually a browser thing. You need an actual user interaction with the page first before you can play videos with audio.

https://developer.mozilla.org/en-US/docs/Web/Media/Guides/Autoplay

question, is it possible to set up cryptography in javascript in a way that even if the key used for encoding is known, it cannot be decoded, only with another private key?

use case is to obfuscate the data I send to the backend, and I don't want it to be so easy to decode as just using the key from the javascript file.

>>105911136
>use case is to obfuscate the data I send to the backend
So literally HTTPS?

>>105911136
What you describe, i.e. "encrypt with one key, decrypt with the second, can't decrypt with the first key" is the textbook definition of asymmetric encryption. Please take or watch a intro level cryptography class before you even think about touching something that actually needs to be secure.

To actually answer your question, yes, if you had a JS implementation of, e.g. RSA, and you had the public key in the browser, the user would not be able to decrypt it, but the backend that has the matching private key could.

This is pointless, however, since the user has total control over the browser and can just intercept the data before it is encrypted by your public key.

>>105911240
Unless he just wants it encrypted to stop sniffers on the network. In which case again literally use fucking https.

>>105911304
I'm assuming anyone bothering to ask about encryption already knows about and is using TLS, but perhaps that's too generous of me

>>105911240
What is the recommended way of doing it in JS? The web crypto functions, stanford library, or something else?

>>105911326
You'd think so but that I legit can't think of any use cases for encrypting on the frontend given what you already mentioned about the user can just get the data from the browser before it reaches encryption state.

>>105911485
What are you even trying to do? Why do you need it encrypted on the frontend? Why couldn't you just send the data to backend through https and encrypt whatever you need encrypted on the backend like everyone else?

>>105911508
>Why couldn't you just send the data to backend through https and encrypt whatever you need encrypted on the backend like everyone else?
I just want to make sure it is encrypted even when https is not usable (internal networks etc).

>>105911588
That's more of a Cyber Security job than a web dev to make sure data across the network is encrypted.

Only encryption you should have to worry about as a web dev is the HTTPS to the server, and passwords and whatever else you want encrypted in the database on the backend.

Obviously there's other cyber sec stuff you have to do as a webdev like sanitation to stop injection, rate limiting, XSS shit, writing for load balancing to prevent against DDOSing, etc and even more if you're also managing your own server like SSH, proper root permissions management, SSL automation, turn off passwords to stop crackers, etc

But encryption over the company network falls into cyber security grounds.

>>105911653
OK, so you don't know if it's possible to do it in JS or not, we got it.

>>105911718
If that's what you took from it instead of you're a retard from trying to do it in js and go check a cyber security general, than that's a you issue. But we could already tell you weren't very smart.

>>105911718
you could try to use use asymmetric encryption, then notice that it's slow, then negotiate a symmetric encryption key with the backend over the insecure channel via diffie hellman, then you'll have a hardcoded public key which you want to rotate periodically, then you want to transparently switch to more secure protocols or key lengths over time etc. what reason could there be to not use https on an internal network? https does not require internet access to work. if this is just a learning exercise, go ahead, but don't trick yourself into thinking that you built something secure

>>105912150
He's a fucking retard, no point in bothering with him.

>>105911136
HEY EVERYONE THIS STUPID ASS NIGGER IS TRYING TO DE-INVENT HTTPS! EVERYONE COME LAUGH AT HIM!

>>105912223

that's why I have alergy to cybersec, imagine a new exploit is born every day, that's their reality

>>105909545
Remember you inserted html characters in to the chat because I was retarded and forgot about sanitation?

Serverless: a meme or not a meme?

Is there such a thing as free VM? Azure costs tens of dollars per month on very minimal use and Oracle seems to charge the same.

>>105914151
It's not going away, an entire OS just for yourself can be overkill
>>105914225
Oracle lets you run 4 CPUs and 24GB of RAM (in one VM or across VMs) for fucking free forever!!

>>105914317

I don't know where to ask this but I am an utter beginner and a noob so:
How do I create my own website? specifically a personal imageboard?
Apart from HTML, PHP and javascript, what else do I need to learn?
Most importantly how do I obtain a domain name? do people really pay for that stuff?

>>105914151

I don't trust it as it encourages vendor-lock, and is a black box for your own purposes. They can charge whatever and say whatever about the numbers on their side and you have to take their word for it. For discrete normal servers you can shop the exact specs and price and know exactly what you're getting. For smaller scale I suspect serverless is fine but if you're a big organization you're going to want to fine tune the numbers, get a dedicated server team and use an appropriate amount of compute power for the lowest price possible.

>>105914349
Idk, try switching to ARM? I just saw https://www.oracle.com/cloud/free/

>>105914438
this is more of a /wdg/ question >>105872088 (OP)

aside from the actual sites code, you need to learn about hosting (managed vs unmanaged), and ideally a webserver (nginx, apache, caddy, etc.). you buy domains for a period of time from a registrar (godaddy, cloudflare, porkbun, namecheap, etc.) and you set dns records for that domain that point to your webserver, which hosts your website.

>>105914589
oh wait im fucking retarded i thought this was /hsg/

>>105914151
not a meme, just poorly marketed by retarded youtubers with a vested interest in pushing a single vendor without disclosing the downsides of serverless

>>105914589
So, I need a host and a domain just to get an empty HTML page online. Okay got it.
However I've seen people host websites for a short period via tor onion links, do these require domain registration too?

How do you guard against CSRF/XSS shit? I have a web app with the following setup:
Frontend in Netlify
Backend in somewhere else (IDK where to decide yet)

So when I send requests to my backend (from frontend) it gets blocked by CORS policy or some shit (this is a simple app, no database just some calculations but I want to learn how to do this safely eventually obviously).

I've been reading some CSRF token implementations but I'm so confused, you set a randomly generated token on the backend (the first time a request comes from your client) and send it back to the client then the client will set the random token on its headers for the next req. You then validate if client and backend has the same tokens to actually/finally process the request (kinda like a 3way handshake, idk)?

2pic4attention

>>105914854
if you want to just host a static site, you can do that with github pages for free and wont need hosting or a domain. you wont be able to interact with a database of serverside software by doing that unless you host an api somewhere else and call to it in the frontend.

no clue about tor stuff, but id imagine they still require servers for them, its just the traffic routing and domain resolution thats decentralized iirc.

>>105914438
>how do I obtain a domain name? do people really pay for that stuff?
yes you buy it, else you might be able to use a free service like once .tk was. Where your shitty.free.domain.com/lol => gibsmedatforfree.tk

>>105914937
As I said, >>105914438 I want to host a small imageboard, somewhat similar to litechan.org (I'm probably going to contact the admin of that site at some point and ask him how he does his stuff).
The empty HTML page was just an example so I know what stuff I need to do before even beginning to write the code for the site.

>>105914999
youll need a host that allows you to run some kind of backend to handle post requests for submitting files/posts (php is fine for this), a database running somewhere (probably also on that host) for storing posts and post->file relations, storage for files, and a webserver to serve the actual site. you can do this all pretty easily with a cheap vps.

How do you create a web app but does not require you to design anything or draw anything? I can't draw or design anything and neither do I have the time to do so but at the same time I can't think of a web app idea that is intermediate level and doesn't require me to design something. I can't think of anything good enough.

can someone explain why removing backdrop-filter on a div punted all of it's contents half a screen to the right

>>105915988
just use bootstrap or a web theme/template

>mfw I remember that people actually create backends in python and javascript

>>105914509
>2x amd cpus fo free
Oracle eats t4g nano for breakfast

I see quite a few web dev jobs using Java or C#. Java seems so heavy to me though. If I start up a Spring Boot app it takes ages (compared to JavaScript, or PHP, or Go, etc). Also the Java language server takes ages compared to other language servers too. Just a fucking heavy language.

>>105916712
yes, and verbose like a mf, full of abstraction, at least spring is

im learning about mvvm, i think i have a somewhat clear picture of how it works, but one thing im not sure about is how an mvvm app handles global services like authentication, without using a dependency injection package an mvvm setup could look like

LoginView(
LoginViewModel(
UserRepository(AuthService, UserProfileService)
)
)


so if i want to authenticate the user on every screen, I would need to make it a constructor dependency on all the views that need it? like

FavoritesView(
FavoritesViewModel(
UserRepository(AuthService, UserProfileService),
FavoritesRepository(FavoritesService)
)
)


?

>>105889588
The Grift Continues

>>105914873
>CSRF
You need protection against CSRF attacks if you use session cookies to handle logins. Attackers could "hijack" that session cookie when the user is logged in and then do some malicious request to the server.

For example: the attacks sends an email to the user and the user clicks on the link in the email while he's logged into the target web app. That click on the link is some request to the server of the target web app (maybe to delete some resource on the users account or transfer something) and this is possible if you only use session cookies.

The best defence against CSRF attacks is to use CSRF-tokens that are stored in the HTML code. CSRF-tokens could be stored in a form in a hidden field or in a meta tag (that you then grab with JS). This CSRF-token has to be randomly generated and sent to the client from the server. It should then be included in every request to the server to verify that it's actually the client (knowingly) making the request. An attacker might be able to hijack the session cookie but he will not be able to hijack the CSRF-token that is in the HTML code of the client (the browser of the real user).

Also, obviously hash (e.g. SHA-256) the session cookies and CSRF-tokens on the backend. Always generate new CSRF-tokens for every form or page (if you put it in the meta tag).

>XSS
You need to sanitise all user input! Always assume that it's malicious if it comes from the client. Use widely used libraries for this. Don't DIY this because you'll probably miss something. Also use CSP and HttpOnly for cookies (JS cannot access them).

Page eight

>>105915988
I'm trying to avoid bootstrap for learning purposes.

>>105917866
this looks like fucking ass

as in this framework is fucking you in the ass

what would you use if interview guy wants you to make a web app in a week with a frontend and a backend if you haven't really used any stack or javascript before

basedom got an update
we have single file components in .html now
reactive, stateful, components in vanilla js and html
no build step
how can it be more based?
https://www.npmjs.com/package/basedom

>>105920716
>4 minutes ago
This is your package isn't it.

That said why would I not just use Alpine for that and do it way easier?

>>105920736
>simpler than alpine
>0 dependencies
>readable editable code
>way more features than alpine
>all the nice parts of modern frameworks without the bloat of .jsx typescript buildtools and shieett

vue is all you need

>>105921307
i don't like .vue files and build steps

I'd say Vite is all you need.

>>105920337
claude code.
I am building an app every single day with a frontend and backend.
you're going to get fired though if you don't know shit though.

>>105920716
Good now add if else and for loop syntax in sfc html

>>105922222
best get I've had in a while

>>105920716
Actually pretty cool, I was looking for something like this for an electron app last week and decided to implement everything without using any npm packages or build steps (jsdoc with type checking) during a vibe session

It's horrendous and for some reason is implemented inside a giant class but hey, it works.

>>105918109
aight ty for the qrd, based anon. Will try to actually implement something this weekend

>>105922224
coming soon

>>105922454
nice work broki
BaseDOM is accepting contributions

>>105922224
otw

>>105916518
I'm learning ASP.NET now and compared to comfy Django I am finding this generally a shitshow

>>105923268
Instagram scaled to 1 million with django

if you don't need to switch, don't

>>105923323
source

Today, I will try a JetBrains IDE for the first time

>>105923467
https://read.engineerscodex.com/p/how-instagram-scaled-to-14-million

https://medium.com/@priyanshu011109/how-instagram-originally-used-django-to-scale-to-millions-of-users-a31bcba39c97

>just google it you lazy retard

>>105923472
best IDEs hands down.

but for frontend I still prefer vscode - just feels more natural to write webshit in that electron bloatware

>>105922526
I'll add this: The CSRF-token should be included in every request to the server if that request changes state, so e.g. POST or DELETE requests. You don't have to include it in a GET request because a GET request should never change state on the server. You should also set SameSite on the session cookies that you use for login. This means that the browser should only send the session cookie if the request originates from the same site, which should protect against CSRF. However, you don't want to only rely on the browser for CSRF protection. You also need CSRF-tokens if you use session cookies to handle logins.

>>105923501
I thought I'd try IntelliJ for Java because I thought it might feel less bloated than VS Code. I just fired up IntelliJ and it's melting my CPU (bear in mind I'm on an old shitty laptop). Damn.

>>105923753
yep that's what java ides do in my short experience, visual studio (the ide) does the same thing fo c#, whatever that other standard kava ide does is the same

>>105917866
if it's really global then just make a public interface that can return an instance of the service and don't bother injecting it

>>105914151
Not a meme. For some situations it's the best approach possible, for some it doesn't make any sense, for most it's as good as any other if it's what you'd like to use.

>>105923753
Sad reality is that in order to be a java dev you have to have a good CPU and at least 16GB memory.

At my current job we run a lot of intensive tests (lots of testcontainers, intesive computations) and my Mac M2 barely handling it.

>>105923532
You don't need this faggotry now with SameSite=Lax

Nuxt 4 is out! Will you guys upgrade? I am using it right now.
https://nuxt.com/blog/v4

>>105923478
>django
>hack (php)
why does facebook pick the weirdest stack

>>105923753
did you try the lightweight option? JetBrains Fleet?

>>105925239
Facebook was originally written by Zuck in the early 2000s right? Back then PHP made a lot of sense for a website like that. Facebook built up a big codebase in PHP, so then they made the HipHop Virtual Machine (HHVM) to make PHP execution faster, rather than rewrite their entire backend in a different language. And the HHVM supports types in PHP too - the dialect of PHP that they named Hack.

As for Django, Instagram was developed outside of Meta wasn't it. Then Meta bought it. So Meta didn't make that initial decision to use Django for Instagram.

>>105925244
No I just downloaded IntelliJ, the reason being that Thymeleaf (HTML templating for Java) was annoying me so I thought I'd try an alternative called JTE. There is an IntelliJ plugin for JTE, so I thought I'd try that.

bros is render (cloud platform) down for y'all? can't get into the dashboard.

>>105924683
You shouldn't only rely on the browser for CSRF protection. The user might be using a browser with some security vulnerability.

why are images drawn onto canvas in firefox crispy and don't have antialiasing

WEBDEV IS SO OBTUSE

>>105925393
ah makes sense i completely forgot they bought instagram ty anon

>>105925648
take a look at what game devs do for a fraction of our paycheck

>>105925648
It's pretty straightforward I think

A few years back I did a bit of work on a textboard specifically for CS resources. It didn't really go anywhere because I didn't have the time, but I've revived it. There's also an IRC server now too.

The hardest part is trying to find people interested in forming a micro-community without spamming it everywhere, which feels very unnatural to me.

https://csbook.club/general/20

:: CS Book Club ::
dev:: ChurchOfTuring
tools:: Typescript/Expess/SQLite
link:: https://csbook.club/
repo:: N/A
progress:: Markdown, mathjax, irc server etc.

>>105925648
web dev is a CUTE

>>105928089
gay post

>>105928079
bookmarked it, probably won't visit it again, but... keep us posted ;^D

>>105928079
>To create an account you need a referral code.
Anon, you want a discord.

>>105928267
I don't because discord is shit. There was anonymous posting at one point but like 4chan it encouraged largely low-quality throwaway posts.

It probably wont work well in terms of getting people to engage but it's just an attempt to encourage more thoughtful posting.

>>105928079
just registered, gonna check it out, you might want to post again at a more american and australian time as well

>>105928079
Looks cool. I am too dumb to discuss CS books and papers though. My CS theory knowledge just isn't very good.

>>105928428
>>105928079
looks like a gold mine so far, might dig a bit in it, thanks for posting anone

should i just suck it up and try learning react or is it worth starting with stuff like svelte? also is it recommended to keep it all TS while im learning? seems like most fullstack jobs are TS across the stack but not all so im not sure

i have like 5yoe doing backend but havent done any front end since college

>>105923268
maybe 10 years ago but now asp.net and c# are filled with syntactic sugar

>>105928843

TS is basically obligatory at this point and is used in all frontend frameworks. Svelte is still a meme that isn't used by employers and React is horrible to learn from a backend background. Try Angular instead.

Can ads be discussed in this general? Does anyone here have any on their websites and what would you recommend? Just want some initial pointers how to start with grifting off idiots

>>105928886
yea last i touched react was maybe 2018 that shit sucked, ill try angular but i still see more jobs for react

>>105928900
Part of the process is trying all the different libraries and frameworks and seeing which of them most closely reflects your views on software development.

>>105872088 (OP)
i wanna make a forum
i made a forum in flask before and i was gonna use fastapi this time but i thought php is probably less gay right, should i install xampp or just install php and mysql myself

>>105929723
thats cool and all but nigga im tryna switch jobs

How do you anons prevent MiTM on android/iOS web apps?
It seems like a lost cause for me given that I can't use my own certs on cloudflare worker (merely fetching some content data and simple logic)
>root/jailbreak check
can be hidden
>vpn check for HttpToolkit
bad due to legitimate uses
>cert pinning
Can't. I am currently on free tier cloudflare so the certs are short lived and pointless to pin.
So what prevents the attacker from trusting mitm certs and/or re routing the request to see the response elsewhere?

>>105930754
For what purpose? Cloudflare already should automatically issue SSL certs. Unless the CA is somehow compromised you shouldn't have anything really to worry about.

>>105930847
The user can trust malicious CA like mitm proxy certs.

>>105930884
afaik at the point of compromising a user's device to install a malicious root cert you can't really do much on your end aside from requiring mTLS (if you really really care about MITM), but no one really does this and the attacker doesn't really need to point to your servers anyways.

If it was a native app obtained from an app store you'd have more control through pinning & mTLS.

>>105930754
the government MitM everyone so ::shrug:: like they have probably a few keys to some of the bigger cert comapnies that are trusted in most browsers. i wonder if they even have the backdoor to letsencrypt.
However, i suppose you could find a way to trip all trusted ca cets on your device (not sure how to do it on mobile OSes) and just put your own ca in the trusted folder.

>>105925244
Jetbrains fleet is still alpha quality. The king monkes banana wars really took a toll on jetbrains

>>105925239
hack/hhvm was good back when it was mostly just a faster php runtime
dropping support for php meant it was doomed to die

Should I minify css then purge unused css in the minifed result, or purge unused css then minify the purged result?

Asking cause I am thinking of forking a project which deals in minification to add purging ability, but wondering which should come first in workflow. Aiming to optimize for size of the css and the speed at which it minifies/purges.

In the end the workflow would be either:
Minify --> Purge Unused --> Compress (GZIP or Brotli)
OR
Purge Unused --> Minify --> Compress (GZIP or Brotli)

Thoughts?

>>105934017
Sounds like you're over complicating things and are going to make your project dependent on specific build plugins that may get brittle to update.

Also be careful about purging styles in the case you use dynamic class names (e.g. string concatenation). Ideally you'd be using an atomic style system such that you won't need to go through a purge step and the compiler does it for you.

>>105934123
The fork is based off of a benchmark which aggregates a bunch of minifiers, if one fails it is simply skipped and the other would be used, also this is only for the prod build step, as a kind of automated squeeze out as much size as possible type thing, you could do it by hand but the idea is that is automatically compares abunch of different tools for you.

As for dynamic classes vs atomic classes, I'm looking to use neither, the project is mainly for classless css projects, css files which apply only to semantic html elements, so the purging should be especially easy.

I'm leaning towards purge then minify, makes more sense to me, what you said about atomic styles made me think that that is probably how they do it, first compile step which also purges, then minification is a seperate file operation.

newbies, DO use chatgpt (or similar) to give you and guide you through js challenges, good stuff

Anyone know what I can build in 2 weeks as a returning beginner (someone who has built beginner web apps before but hasn't done so in a year) for a college project? My goal is to spend 10 to 12 hours a day for the next two weeks, but in reality I'll probably be only spending only 5 to 6 hours because I'm a lazy shit. I also have to use github to update my project regularly and has to be publicly available but idk how that will work, I never really used github to upload and update a project, just to git stuff off of it. I have to build something that is decent because this project is for the class above the beginner level so it can't be a simple web app anymore and the app has to be for commercial use. I don't really know whats good enough but at the same time easy enough to where I can complete it and write a technical report on it in 2 weeks. I just want to at least say I've tried even though it might be too late.

>>105935154
your current knowledge + a bit of google and chatgpt will go a long way, consider paying for a month, in fact consider claude code entierlly, paid that is

>>105935632
but AI is completely prohibited in the instructions for my project. Also I'm pretty sure colleges have programs that detect chatgpt. I once wrote some code from a YouTube video and somehow it got tracked to that video. Also its bad to use AI for something you're not really an expert on because you won't really be able to notice small mistakes an AI would typically make but someone who corrects hundreds of projects will.

>>105935809
I see your point, then just pay for a course that can hand hold u in udemy

How the fuck do the web interns at my workplace manage to vibe code a bunch of new features this easily?
There are 10,000 codes files in there, so how did they break through the word limit?

>>105889541
only use non-US companies so you can't be v&

>>105936750
What are you planning to host which would be illegal in the US but not illegal in Germany?

Then again, do I really want to know? Perhaps I don't.

>>105882340
>What difference does it make if your node/express takes 15ms to handle a request compared to C++'s 10ns of manualy writing HTTP/1.1 requests?

a few hundred megabytes of memory at a minimum

>>105914873
she's more hair on her moustache than her armpit

>>105889541
digitalocean is less cancer imo. Hetzner hustles you hard and then drops a fucking bomb on you out of nowhere. I haven't had any issues with digitalocean other than it being more expensive and being geographically closer to me.

>>105889588
eh. AWS works when you need to solo deploy a functional high availability service with "infinite" scale and "just works" backup solutions.

If you have competent IT or good control over your infrastructure, it's a complete meme that will be a cost blackhole.

>>105872088 (OP)
htmx is a meme, it becomes apparent if you try it for longer than 1 day instead of just mindlessly shilling it

>>105882260
only if you have atleast 6 years expierence in c++

>>105936947
Good take. I've never tried using it but I just don't see the point in it when I already know JS/TS to some degree. Maybe HTMX would be appealing for people who don't want to learn JS/TS though.

>>105896748
because SQL still sucks but retards are psyop'd to make hacky shit on it.

>>105936973

NoSQL was curbstomped. It's easier to get started with and many tried, if it was better it would have won easily, and couldn't. It's highly limited vs. the SQL model.

>>105936912
>Hetzner hustles you hard and then drops a fucking bomb on you out of nowhere
How? It's cheaper

>>105936986
no it wasn't retard.
every major DBMS has carve-outs for document oriented storage.
Postgres, SQL Server, Oracle, MySQL, even fucking sqlite3.

production data-models all look like grotesque beasts of synthetic keys, denormalized data relying on triggers to clean it up, etc. SQL is a shitty meme.

>reading thousands of javascript codes in one script tag
Fucking kill me

>>105935154
how about a student chat app or a teaching poll. both can use websocket so you can update stuff live directly. if uni auth is not possible you can just have it be anonymous and ensure device fingerprinting.

>>105935154
>in reality I'll probably be only spending only 5 to 6 hours because I'm a lazy shit
5-6 hours of intense focus are what most people can do at most. the rest of 6 hours will be just wasted on shitty code that you will have to fix later on

>>105937010

Without fail there's always some autistic narcissist in software that thinks he knows better than the most war-weathered unstoppable technology out there. Literally nothing in software has been as evergreen and unstoppable as SQL.

>>105937064
>he says ignoring all the major commercial """SQL""" DBMSs and Schemas are basically NoSQL, document store pigs with lipstick
k kiddo.

>>105937092

The fact that SQL engines added an extra feature to use document oriented storage doesn't mean they shifted to that model. It means they immediately made NoSQL redundant since you can do the same things with their engine as well.

Document store is horrific at scale. Your queries will be slow dogshit. You pay for the convenience with lack of optimization.

>>105937064
>>105937010
nta, relational is good at relational non relational is good at non relational, "simple" as
t. I'd rather take opinions of DBMS from my DB teacher with actual professional enterprise experience than "pro" butthurt fighters on /wdg/

>>105937224

"non relational" is just a key/value store. That's never been a novel thing. That's why relational is your first consideration. NoSQL didn't pitch itself as an alternative, it pitched itself as a SQL killer, because we already had k/v and other forms of simple storage. Mongo wanted to optimize JSON docs to operate as effectively as SQL objects and it simply didn't pan out.

>>105937224
>nta, relational is good
until it isn't and you start pulling out gigacopes to cover it up, ORMs being the first comment in this whole reply chain being a good start of the cope.

>>105937254
>didn't pitch itself
what does marketing have to do with real world tech use cases and performance?
>>105937278
use cases, simple as, you can go full retard with both relational and no relational, in fact going full retard is the easiest path
you didn't talk to my DB teacher about it, so you don't have actual pro insight, you have more like poo insight

>>105937298
>you didn't talk to my DB teacher about it, so you don't have actual pro insight, you have more like poo insight
meant for:
>>105937254
but actually applies to both posters

pro-tip: I hecking love json isn't a valid reason

>>105937315
The valid reason is having control and not being limited by the giga shit query language.
I still have to stop retards making schemas that make performant pagination neigh impossible. You don't realize just how bad it really is in the real world. Having a B-Tree that has transactions and is crash tolerant is already a massive plus over any shitty DBMS for those that know. If all you're doing is shuffling memory like a cuck and don't care if your queries cause page loads to be 10 seconds or have "reporting" take out your database because of how bad they are then it doesn't matter.

>>105937369
actually implementing NoSQL well is way harder than implementing SQL well

>>105937378
>>105937369
Just wish my teacher could shitpost with us about it, but no, he said NoSQL is more performant with huge amounts of data, but it came with the cost of "it gets very complicated", this nigga was a DB genius when compared to us plebs

>>105937402

It is not, which is why no one uses it "for huge amounts of data." He's referring to when people use DocDbs as key/value stores, not when they try to use the values within relationally. Nothing beats a key/value store for performance which is why it's also used in the caching model. If you have your id that's all you need. But if you start querying data doc model falls apart quickly.

>>105937426
I'm just saying if NoSQL is so bomb why Amazon was using it exclusively to track orders and nothing else?

>>105937451
>>105937426
"what dbms does amazon use and for what purpose each?" read carefully and you'll see

Jetbrains? More like Jeetbrains.

>>105937810
>implying jeets get handed licenses which cost money
they're probably using notepad++

>>105937451
DynamoDB is nosql though...

>>105937879
that was part of my point pham, don't stop at that, keep reading

>>105937951
oh. fair I guess.
sql seems overblown to me honestly. graph databases in the form of triple stores seem a lot better at enforcing real world constraints.
think like OrientDB. sadly it's java jeet ware, but whatever. No one stops me from making my own triple stores on-top of RocksDB.

I just hate how SQL is structurally wrong, non-composable, can't be formatted in any sane way, ends up becoming a performance cliff once you try and ask more questions in a query, etc. I just don't understand why this boomer shitware is still popular outside of sheer momentum and babby duck.

>>105938061
it's more data-complexity friendly, same complexity (extremely relational) deal with data would be clown world to implement in NoSQL, as I said, use cases pham

>>105938108
>extremely relational
graph databases are superior here

>>105938111
anon... relational data is a shit show in NoSQL, doable, but a shit show, you might get BETTER performance in NoSQL, despite that you'll have to fucking dig data structures and duplicate data out the ass, imagine having to maintain that frankestain over time, I wouldn't want to

>>105938141
>having actual relations is a shitshow
wtf are you talking about?

>>105938177
relational data in NoSQL means duplicated data, or in DB theory lingo, data integrity, which you'lll have to sacrifice and compensate with extra labor in NoSQL, and is a given in relational SQL

>>105938204
no, it doesn't.
https://orientdb.dev/docs/3.2.x/datamodeling/Concepts.html#relationships

educate yourself retard.

>>105938208
>using NoSQL relationally
see, that's precisely what the Amazon's NoSQL (dynamoDB) told the audience to NOT DO, because it tanks performance, see now?

>>105938220
* the Amazon's lead DynamoDB dude told the audience in his talk

>>105872250
I'm working on a new imageboard engine, but the one I currently run has a board with files being optional in threads.
people enjoyed it, and there was a real risk of the main board dying so I had to cuck it
>OMG it's just like a chatroom!
also feet

>implement new payment processor for a specific geo region
>works perfectly locally
>keeps breaking on staging for weeks
>ci keeps breaking because of others
>upstream vendor we depend on for data keeps breaking
>Product is pissed
>AE is pissed
>clients are pissed and some are close to churning
>all because people promise customers unicorns while our system is an overengineered piece of shit where everything breaks if you make a small change
I'm tired my dudes

>>105938235
>>105938220
>>105938208
I'm not sure but I believe this might be the talk I'm talking you about:
https://youtu.be/HaEPXoXVf2k?feature=shared&t=2921

>>105938455
everyone wants to go back end until is time to do real back end shit

>>105938561
We created an abomination which is quickly growing out of control. Some of us tried to warn them, but they wouldn't listen.

>>105936868
and sec+maintenance nightmares

>>105938626
sure... I'd say use Rust instead since C++ is kind of shit and you won't find good composable async (co_await) friendly anything.

if i go live with my website will people bully me for my code

>>105938586
surely, you'll know the position of the golden boy, past collaborator as architect comes back this time as new application new version design and implementation

>>105938670
No.

>>105938810
oh ok because it's really fucking bad code

>>105914151
Classic old school PHP web hosting was serveless but good and not jewish

>>105914151
giga meme.

>>105938822
Ok. You won't learn anything if you don't deploy it or share it.

HTMX
more like
HTMSEX (Gay)

>>105938670
Well nobody will see your back-end code unless you make it public on GitHub or whatever. As for your front-end code, just minify it.

i thought it would be cool to see digits across all boards, thoughts? (not live yet)

>>105940231
neat, that'd be like candy for the script kiddies though

>>105940280
>>105940231
and by that I mean, that the first to ever figure out how to get gets excluding the other getters, can derail your whole chan/board

>>105940291
>>105940280
its moreso meant for just seeing digits, not intended to function as a chan. its already trivial to bot digits anyways

>>105940387
I see, very neat
>its already trivial to bot digits anyways
are any of them auyto solving captchas to do it, vs just buying an account I mean?

>>105889249
How did it go?

>>105940562
probably went awfully, none of us wished him luck

>>105914151
It's the biggest meme in tech rn because everything runs on servers. "Serverless" is just an abstraction that abstracts away the server. You're paying for that abstraction so it's going to cost you more and there's information asymmetry. Meaning you, as the customer, have less information about what it should cost. That means that the serverless providers (AWS, Azure etc.) can overcharge you even more.

>>105928079
>>105928267
I updated my site so people can now post anonymously. Registrations still need a referral but at least there's no barrier.

https://csbook.club/off-topic

>>105936428

>>105936428

a

just listened to 6 hours of this dhh nigga on lex friedberg podcast. hes right web dev is simple and solved. we should focus on the things that matter and therefore I will stop using cloud and start using metal and ruby on rails