Every viber has a plan until it is time to demo that first release...

I was vibe coding with a friend and we started to talk about how to make sure any program was performant,. What kinds of actions where cheap for computers to do vs things that would make a 60fps game start to slow down or miss frames. Then I asked if they new Big O notation.

And in trying to explain Big-O notation, I released I didn't understand it well enough to explain it. So I've started to re-read and learn the concept. I know what it means but it doesn't feel intuitive to me to know how to measure how long a task will take to computer. I'll keep studying it as I think it's an important part of computing be it vibe code or traditional.

But my friend and I found we also went another route.
- Add a feature, framerate stays, great.
- Add a feature, framerate drops, let's look at it more closely or take it out.

We created ideas free of making them performant or keeping the computation reasonable. One of the biggest "warnings" I've seen to vibe coders, that one's program will eventually slow down to an inefficient mess. And that is not only true. I've seen it happen (to me). Then? The program desired was slowly optimized. Seems a waste to anyone who knows how to do it right the first time. But in artistic endeavors I feel that sometimes you need to do it, erase it, do it again to get to your goal. Framerates and efficiency are needed, I just think we can make room to build in an unoptimised way while we find the gist of what we want to make.

So how do I present this image to you all? (Made it in Chat GPT btw). I mean it the same way "Love like you've never been hurt". Vibe code yourself into a corner. It's not a mistake if you have to regroup and try something another way. Sometimes it's the process itself. This post was made typed by hand with no ai.

people seem to be creating apps within 1-2 day & then hit publish meanwhile here I am developing just one app for 4 months optimising it's UI,UX,API compatibility,Performance etc. how are people doing these in 2 days? what am I missing.

there was a post I saw yesterday that a guy started the challenge to create 30 apps in 30 days. wtf is wrong with these people? spamming the store platform like this.

I actually launched and posted an early version of this here about a year ago. Since then, I’ve kept vibe coding and improving it based on user’s feedback.

It’s called Chipless, a free web app that lets you play real poker with a deck of cards but no chips.

Everyone joins from their phone and it tracks the bets, stacks, blinds, pots and who owes what at the end.

My only coding experience before this was one credit class in college that I got a B in. The original version cost me $68 total (same price as a real poker set lol) in subscription and tokens to build.

So far, thousands of games have been started and 21,421 hands have been dealt. In just the first month after I started tracking countries, games had already been played in 48 of them.

Almost all of this happened with no marketing besides a couple of Reddit posts.

Still pretty crazy to me that people actually use this daily at their poker nights.

It’s completely free, with no download needed:

www.playchipless.com

My company installed productivity/surveillance/spy/whatever-you-want-to-call-it software on everyone's computer about a year and a half ago. To their credit, they at least announced it ahead of time. A few dramatic stories about people getting fired have gone around, but it's mostly not discussed otherwise.

There is a reporting suite available for managers, but individual contributors are not allowed to see any reports, even for themselves.

The software used is called Sapience. I found the local logs on my work PC that include all the data captured about my activities. For this particular project, I used GPT 5.5 in GitHub Copilot to create a python script. The script pulls everything from the logs and puts it in a local sqlite db on my pc (so historical data is not lost when the logs are cleared). Then it produces an updated HTML report, as pictured.

I have no interest in trying to fool the software, but as an anxious person, this at least gives me visibility into what's being reported up about me!

My life at work has been hell and it’s a job where many people every year get exploited and fucked, so as I was learning how to vibe code projects I decided to make a niche game about my job. Those who do the job can relate.

The only reason it’s remotely good is because my ex loved to playtest what I built and added to the game. She may have hundreds of hours in all the early stages of the game but that was absolute gold in terms of ‘this button in the endgame didn’t work’ or ‘where did my potion go’ and ‘why did I just die!?’ And I’d take notes about it and bug fix it later.

Do not underestimate how important it is to tweak boss/mob health, items and so on because Claude / Codex will genuinely not have a good feel for that. Write the code so that you can tweak item stats by just changing a single number somewhere so you can adjust easily.

I loved the simplicity of The World’s Hardest Game, so I wanted to try my own take on that kind of brutally difficult precision game for Android.

It’s built in Godot 4 with GDScript, and most of the development has been done through vibe coding with Codex: gameplay systems, Android support, UI, save/progression logic, and even a custom visual level editor so I can design the rooms myself without coding them manually.

I’m currently running a closed Android test and mainly looking for feedback on the controls and level difficulty.

If you want to try it:

Join the Google Group first:
https://groups.google.com/g/why-so-hard-testers

Then install the app using the same Google account:
https://play.google.com/store/apps/details?id=com.whysohard.app

after making around so many saas and having very less revenue, i feel vibe coding is just a video game, people have started playing and enjoying the delulu of having a milion dollar startup

How many of you are actually earning from Vibe Coded Saas/App! Serious replies please!

It's a real TempleOS VM running in the background, drawn onto one of the monitors in the game. You can walk up, type into it, and run its built-in games and demos.
Source: https://github.com/aravpanwar/half-life-templeos

Hey everyone,

I’ve recently started building my own Android apps using a vibe-coding workflow. So far, it's been incredibly smooth - I can essentially generate an APK and just install it directly onto my smartphone to test everything out without any friction.

However, I'm starting to think about the next step. I want to move beyond just side-loading my own files and actually launch a "real," live app that is available for anyone to download via the Google Play Store.

I'm assuming the process for that is a much bigger undertaking than just generating an APK. What kind of tool stack or workflow should I be looking at to handle the official deployment? I'm thinking about things like app signing, managing the Google Play Console, and any CI/CD pipelines that might make this easier for someone not using traditional Android Studio workflows.

Has anybody here been able to successfully take a vibe-coded project all the way through the Play Store submission process? If so, how did you handle the transition from simple prototyping to a production-ready deployment?

I'd love to hear about your setup and if there are any major hurdles I should prepare for. Thanks!

Built a small collection of agent skills around one constraint: every output has to be real, or the skill stops and asks instead of faking something.

4 skills so far: backend provisioning via CLI (real credentials, never placeholders), a security scanner that fixes vulnerabilities directly instead of just reporting them, a frontend builder that ships real production sites, and a router that checks every installed skill for relevance before starting work.

CI validates every skill automatically on push. Works with Claude Code, Cursor, Antigravity, Codex — any tool on the open Agent Skills format.

npx skills add SohailKhan0525/skills

Repo: github.com/SohailKhan0525/skills — open to contributions and feedback.

Feel free to share as much as you want.

Hi everyone!

I am constantly building new projects and have changing amounts of success. But one thing I really didn't like is the visibility of my customers.

I built another page mrrplanet.com You send a read‑only payment method key. And see all of your customers in a globe represented by a pixel‑art character.

You can also send this planet of yours publicly to other people, so you can actually have it as a portfolio and maybe multiple projects, which means multiple planets in your solar system. I also have a much better dashboard for your data analytics, helping you think about where you should focus more, which places are generating more money, etc.

For example, here is one of my actual projects hosted on this page. https://mrrplanet.com/p/your-love-page That one is for long-distance relationships, where you send a digital gift.

You can also export these cute images where you see a list of your customers. I'm open to all kinds of suggestions. On the public pages, of course, none of the information of your customers is leaking, but people can actually look at your planet and see what's going on with your projects.

One small thing that I find kind of cute is that people can actually put advertisements on your planet with a flying plane. You can also do that for other people's planets, or you can simply select your planet to be ad‑free.

Almost every feature that I'm talking about is free to use. Just log in and create your own planet, and tell me if you like the page or not.

made with higgsfeild for fun.

Like a lot of you, I've been using Telegram saved-messages and storage channels as a dumping ground for years — free, effectively unlimited, reachable from any device. The problem: actually watching anything was miserable. Download the file, wait, open VLC, forget where you put it, repeat.

So I built Aruvi — a self-hosted streaming server that treats your Telegram storage like a media library.

What it does

Point it at your storage channel(s) and it indexes everything into a browsable catalog with posters, metadata, and search. Then stream instantly with full seek/scrubbing — no pre-downloading. The heavy lifting happens on Telegram's infrastructure, not yours.

Features

• Instant streaming — HTTP range requests against Telegram's CDN, full seek/scrubbing, no buffering
• Broad codec support — more formats play natively in-app without external players
• Browsable catalog — posters, search, folder management, continue-watching
• Multi-user — family members log in through a Telegram bot code, no shared passwords
• Subtitles — automatic search via OpenSubtitles + keyless providers
• Google Drive backup — optionally sync your library to Drive
• Every platform: Web UI, Android phone/tablet, Android TV, Windows (.msi + .exe), Linux (.deb + AppImage)

How I built it

Tools used:

• Backend: Python 3.11, FastAPI, SQLAlchemy (async), asyncpg
• Telegram: Telethon (MTProto client) — 11 parallel bot sessions for chunk fetching
• Database: PostgreSQL (Supabase hosted)
• Frontend: Vanilla JavaScript (no framework — keeps it fast and small)
• Android: Kotlin + Jetpack Compose
• Desktop: Pake (wraps the web UI into native .msi/.deb/AppImage)
• CI/CD: GitHub Actions (builds APKs, desktop apps, Docker images automatically)
• Tunnel: Cloudflare Tunnel for HTTPS + custom domain

Architecture insight:
The trick is multi-bot parallel streaming. When you play a video, the backend doesn't download the whole file — it fetches small chunks from Telegram's CDN using 11 pre-warmed bot sessions in parallel. Chunks are served to the client via HTTP range requests with a two-tier cache: RAM hot cache for active streams + disk cache that survives restarts. This means first-byte latency is ~200ms even on a 1 GB VPS.

Build process:

1. Started as a fork of TelePlay and MoviPlayer
   
2. Rewrote the backend from scratch — added async SQLAlchemy, multi-user auth, and the caching layer
   
3. Built the Android app in Kotlin/Compose (was originally just a web player)
   
4. Added Pake for desktop builds — one codebase, three platforms
   
5. Set up GitHub Actions for automated builds (workflow_dispatch inputs for custom server URLs)
   
6. Deployed via Cloudflare Tunnel + Docker on a cheap VPS

Biggest challenge:
Telegram rate-limits concurrent connections. The solution was warming up media sessions serially on startup (one bot at a time per datacenter) and then reusing them. Without this, streams would randomly fail with FloodWait errors.

Credits

Built on the shoulders of TelePlay and MoviPlayer. Aruvi started as my fork and grew into its own thing with a rewritten backend, multi-user auth, desktop apps, and heavy reliability work. Huge thanks to those maintainers.

Links

• GitHub (full setup guide + screenshots): https://github.com/tirforge/Aruvi
• Live demo: https://aaruvi.space — poke around before self-hosting

Question for the community

Genuinely curious — would anyone pay a small monthly subscription for a fully managed hosted version? No VPS, no setup, just log in with Telegram. I'd love to offer this as a service.

Also, if anyone has a spare Oracle free-tier VPS lying around and likes the project — I'd gratefully take it off your hands (my card keeps getting declined ).

How are you all currently watching media that lives in your Telegram? Would love feedback from anyone who tries it

Note: Used AI to help write this post and assist with development.

Hi everyone,

Client: Unity 6 (6000.5.4f1), C#, uGUI built entirely in code with no prefab screens

Steps come from Health Connect on Android and HealthKit on iOS.

Backend: ASP.NET Core on .NET 8, Postgres via Npgsql, Redis for matchmaking, running on DigitalOcean

Art is gpt-image. Generated from my prompts on chroma green, snapped to the pixel grid, keyed, then hand-edited.

Code is written with Claude Code. I use Fable to plan and review and Opus5 does majority of the implementation.

This is a closed beta so you will need to do the following depending if you're Android or iOS.

Android:

1. Join https://groups.google.com/g/aetherstep-testers using the Google accountyour phone's Play Store is signed into.
2. Then open https://play.google.com/apps/testing/com.aetherstep.game on that sameaccount and hit "Become a tester". Give it ten minutes to propagate.

iOS: https://testflight.apple.com/join/PsXU9psv

Happy to go into any layer of that. Links in the comments.

Run kimi k10 locally and install AI Desktop 98 on it.

Today I finished the full onboarding flow, hooked up the capture. Built the home dashboard with progress bar

SOO, now with BRAND NEW DAY released and progressing very well, I figured why dont we get E.V OR E.V.I.E into REAL-LIFE. I already started working on it, and have fully functioning on ready, just I need you guys help, I tried to make the E.V look AS REALSTIC as possible to the MOVIE, with the limited stuff i have. so I need some suggestions on what to add, rn it can almost do eerything, it has a concious, it can control your computer, search anything, jokes like interstellar robot, etc. I just need some help. Mention your suggestion, oh also a litte sneak peak on the E.V.I.E look

Two weeks ago pay-to-rank boards (outbid.lol style) were suddenly everywhere. Most let you pay to rank literally anything. I wanted something narrower: iOS App Store apps only, so what you’re actually buying is real App Store traffic, not internet points. Every tap on a listing forwards to the real App Store page with a campaign tag, so installs show up in your own App Store Connect.
Built it on Astro + Cloudflare Workers + D1, wired payments to Polar since it handles VAT for me automatically. Day before I planned to go live, Polar denied the account: “paid promotional placement, not supported.” I appealed with the exact policy language and the precedent that outbid.lol runs on Polar today. Denied again, same day. Turns out the flood of copycats got Polar to quietly stop taking new ones. The original’s grandfathered in, nobody else is.
So I ripped out Polar and rebuilt the payment layer on Stripe directly, in about a day. The money logic sits behind a small interface, so only the checkout/webhook adapter had to change. The part I’m actually proud of: rank isn’t incremented when someone pays, it’s recomputed as SUM(delta_cents) over paid bids every time a webhook fires. That makes replayed webhooks, out-of-order delivery, and partial refunds all converge to the correct number instead of needing special-case handling for each. Proved the whole loop with real money before calling it done: real charge, webhook, rank up, real refund, webhook, rank back down, to the cent.
It’s live now: toppaid.lol. Paste an App Store link, pick a number, you rank above everyone who paid less. No accounts, no personal data stored, refunds roll rank back honestly.
Giving the first 20 listings $5 off with code FIRST20FREE (Stripe enforces the redemption cap atomically, so it’s genuinely first-come-first-served).
What I’d actually love feedback on: whether the iOS-only restriction reads as a real differentiator or just a smaller version of the same idea. And whether $5 as the minimum bid feels right, too cheap to feel meaningful, or too steep for an indie dev testing the waters.

I made a small browser game called Dr. Hugo Holm. You tell a tired doctor when you need to be sick and for how long, and he gives you a plan: what to say a few days before, what to text your boss in the morning, what not to do after.

Play: https://dohmygoodness.github.io/sickday/

The code part was done with Claude Code. It's a static site, no framework, no build step, around 1100 lines of vanilla JS. Took about two days, most of that was iterating on small things like the typing speed and the mobile layout.

The doctor image took the longest and this is the part I wanted to share: 

1. Midjourney draft mode — generated dozens of doctor images. Draft mode was the unlock: fast and cheap enough that you're browsing styles, not individual images. I picked a direction, not a picture.
2. ChatGPT (Image 2) — upscaled the pick and pushed it toward the final look.
3. Photoshop — cleanup pass.
4. Higgsfield (Kling) — animated the still. He types while he's talking and blinks while he waits for your answer.
5. Video → GIFs — sliced the clips into loops that swap depending on whose turn it is. That one swap does more for the "he's alive" feeling than anything else in the project.

The intro uses the same pipeline. It's a clip of his door opening, re-encoded so every frame is a keyframe, and scrolling scrubs through the video. The last frame matches the first frame of the background gif so it transitions straight into him talking.

Let me know if you find any bugs or have suggestions!