pablo's spot’s Post

#Day147 of Sheryians Coding School 🚀 Small visual experiments often teach the most about physics and interaction. Today was about exploring particle behavior and understanding how tiny logic changes can completely transform motion. 1️⃣Self Task: Particle Push Effect Experimented with interactive particle behavior Initially implemented attraction between particles Flipped the force sign to create a repelling (push) effect Observed how minor mathematical changes drastically affect motion A fun experiment that can be reused later for interactive UI effects Source Code: https://lnkd.in/dKvJF3jS Live: https://lnkd.in/dgqg6TZQ 2️⃣DSA: Leetcode Problems Practiced Pow(x, n) Running Sum of 1D Array 3️⃣Self Study: System Design - Monolithic Architecture Also called a Centralized System Core parts: frontend, backend, database Entire codebase exists in a single repository Components are tightly coupled Advantages Easy to implement and understand Faster to start for small/basic applications Disadvantages Tight coupling makes scaling and changes harder Maintenance becomes difficult as the system grows Also clarified the difference between: Website → mostly static content Web Application → dynamic, interactive content 4️⃣Live: Backend day 16 Integrated ImageKit into the mini-project Learned why image compression is important (performance + bandwidth) Explored best image formats for serving optimized assets 🌟Key Highlights Explored physics-based interaction effects with particles Strengthened mathematical intuition behind motion logic Continued building System Design fundamentals Learned practical backend optimizations for media handling #Day147 #SheryiansCodingSchool #FrontendDevelopment #Canvas #ParticleSystem #SystemDesign #Backend #DSA #LeetCode #CodingJourney

That clean working code you see… Behind it there were: 20 errors 15 Google searches and at least 3 moments of “yeh kya ho raha hai” Coding looks easy only after it works. 😄 #CodingLife #StudentDeveloper

🧘♂️ The Zen of Coding: Embracing the "Blank Page" We often feel a strange pressure in tech to always be adding. Adding features, adding lines, adding complexity to show we’ve been busy. But true mastery often looks like a blank page and a quiet mind. The Zen of Coding teaches us that the most elegant solution is often the one that removes the most noise. Have you ever spent four hours researching a complex architectural pattern, only to realize you didn't need it at all? That realization isn't a "waste of time." It’s the ultimate win. You saved your system from unnecessary weight and your team from future technical debt. The Lesson: Don't be afraid of the silence between commits. Some of your most productive work happens when your hands aren't even on the keyboard, but your mind is simplifying the problem until the solution feels effortless. When was the last time you decided not to build something because a simpler way appeared? Does "less is more" resonate with your workflow, or do you prefer to build robustly from the start? Let’s find some balance in the comments! 👇 #ZenOfCoding #SoftwareEngineering #Simplicity #MindfulTech #Python #CleanArchitecture

Vibe coding just wasn't cutting it for me. Developers think in code, not paragraphs! So I invented Shadow Coding! 🚀 Compared to vibe coding, Shadow Coding gives a higher-quality output using fewer input tokens, all while using a free open-source model! 🤯 How does it achieve this? Easy... Shadow Coding let's you prompt with code instead of plain English! Check it out: - Watch The YouTube Video: https://lnkd.in/dkZBwd4i - GitHub Link: https://lnkd.in/db5UJVNA #ai #vibecoding #coding #claude #claudecode #vscode #programming #aicoding #cursor #dev

Coding can be very humbling. You start confident. Then three transcripts in, everything feels messy. You create codes. Then you merge them. Then you delete half of them. I’ve learned that if coding feels too easy, I’m probably not looking closely enough.

I want to share my POV on vibe coding. So recently, I was curious about why so many people enjoy vibe coding, so I decided to try it myself by building a full-stack application with the help of AI. Here’s what I learned: • Vibe coding is great for accelerating learning, especially if you already have some experience. ( In my case, I learned a lot more about Python during the development process. ) • To avoid an “AI-designed template” look, I believe we should still have our own design direction (from Figma or even simple sketches). This is optional, but it helps keep the product intentional. • We must make sure we actually understand what has been built. Otherwise, it becomes a time bomb for the future. At one point, AI deleted my migration files and created unnecessary ones or if something breaks or behaves unexpectedly, we need to be able to trace the source of the code ourselves. • AI is also useful as an idea generator, especially when we feel stuck or uncertain about the next feature to build. My takeaway: Vibe coding works best when AI is a partner, not the driver. You still need engineering judgment, product thinking, and ownership of the code. https://lnkd.in/dprRPKvY That’s my perspective on vibe coding — we might have different opinions, and I’d love to hear yours. What’s your take? 👇 #VibeCoding #AI #RAG #AIFlow #MyJourney #SoftwareEngineering

Why learning syntax is the least important part of coding Novice programmers often obsess over the syntax of a specific language, but Sebesta argues that true mastery comes from understanding the underlying design philosophy. Instead of just memorizing keywords, you should analyze the trade-offs designers made between readability, writability, and reliability. For instance, understanding why a language restricts pointer arithmetic reveals its stance on safety versus efficiency. By studying these architectural decisions, you stop seeing languages as isolated tools and start seeing them as variations of the same fundamental concepts. This perspective allows you to learn new languages in days rather than months, effectively 'future-proofing' your career. —Concepts of Programming Languages https://lnkd.in/gADmfnge #BookReview #Eblite #ProgrammingLanguages #ComputerScience #LanguageDesign #ProgrammingParadigms

Claude Code Is Awesome! But Do We Still Need to Write Code? The question isn’t if AI will change how we write code. It’s whether writing code will even matter anymore. My bet eventually, humans won’t write code at all, just as we once outgrew low-level programming. Check out essay I wrote recently: https://lnkd.in/gKTFc82b

Vibe coding has killed Show HN. The projects themselves didn't make Show HN great. It was the discussions about the decisions and *how* the problem was solved. By offloading problem solving to an LLM, that depth is gone. Software engineering isn't about writing code. It's about solving problems and understanding *why* a solution works. More on this here. I highly recommend the article linked below.

Last week, I got the opportunity to co-host a workshop representing Cursor at UW Madison with Ojal Prasad. We spent the session diving deep into how AI-native coding is changing the game for developers. The room saw how Cursor isn't just a multi-file editor, but also an autonomous agent. We demonstrated how to set a high-level goal and let the agent research the codebase, generate an implementation plan, and then execute across multiple files while maintaining type safety.