Hey again everyone! Today I want to tell you how I got interested in web development, what and how I studied, what difficulties I faced along the way, and what came out of it in the end.
Backstory
I don't know if you know this or not, but I’ve always been drawn to the Web. The Internet seemed mystical to me, untamed, a place where everything is stored — a true treasure trove of human knowledge. Basically, something like The Wired from Serial Experiments Lain — not completely abstract or metaphorical, but still with a layer of mystery. At least in my imagination, it looked like that.
Unfortunately, in reality, it's just cables laid on the ocean floor connecting servers to each other. But let’s pretend it’s magic ಥ‿ಥ
I always liked the Internet for its "local feel" and the ability to find anything anytime. But even more — for the memes. So one of my already fulfilled wishes was to make something online. It didn’t matter how good or bad — the main thing was that it existed. That it made electricity burn in datacenters. (•̀ᴗ•́)
First Website (or How Not to Do It)
As you might guess (or as many people know), I had never done web programming in my life. I’m kind of an ML engineer — what do I have to do with websites? So the idea of making my own site was buried deep in the back of my mind. My only connection to web was through Gradio and Streamlit.
But then some programming courses started at the university — and that kicked everything off.
The Very First One
The first site was just a regular web lab assignment. Built with Django, with a primitive design and minimal functionality — we had to make a simple blog.
By the way, even back then the idea inspired me, and I wanted to make my own real blog (more on that later). The site was primitive in everything — from the database to the CSS. But it worked. Because everything was done by following the instructions. It’s dead now — Anaconda isn’t too keen on hosting stuff like that. You get the idea.
Overall, nothing great, but for a first time touching HTML and CSS — I think it was more than fine. Or maybe not, you decide ʕ•ᴥ•ʔ
The First One (on Vibes)
I gained some experience. And quickly forgot it.
Did anything nice come out? Not really. And it all took a lot of time. I think I would’ve quit if not for a new idea — to make a university schedule site.
Because besides a PDF, there was nothing. And I wanted to be able to just send a link, open it on my phone, and everything would work. Like in Tomsk, for example.
Also, I found out I could use this project to close my internship credit — motivation shot up.
I built the project almost entirely on “vibecode” — basically by feel. About 98% of the generation was done by Claude Sonnet 3.5, the rest I edited manually, but mostly I was just a monkey waiting for stimulation in the form of a working feature.
Let me say right away: I don't recommend this method to anyone.
After about a thousand lines, the neural net loses context and doesn’t know what’s going on in your code anymore — and neither do you.
And most importantly — you don’t actually learn anything 😊
At some point I had to do dozens of generations just to implement a very small and simple feature, and what happened during debugging... I’d rather not say.
But in the end, it worked (miraculously).
It’s still live on GitHub Pages: unitime
I was planning to add OCR to parse the schedule PDFs. But then I realized — I didn’t need it anymore. I did what I wanted — and that was enough. And that was the right call.
There’s no point supporting neural junk if you can just sit down and do it properly.
The Very First (Proper One)
That’s when I really felt it was real. My thoughts had taken shape. Wow.
But a couple days later, once the euphoria wore off, I looked at the schedule site again and realized something:
Nah, AI still won’t replace us anytime soon ಠ‿ಠ
And so an old dream came back — to build a site I wasn’t ashamed of. And most importantly — to build it myself.
I started purposefully learning React. And Next.js in parallel. I always knew it existed, but I’d never touched it.
Big mistake.
Because Next is the best thing to happen to web dev. Incredibly convenient framework with top-notch documentation. Everything is explained in detail.
I decided to make my site minimalist. Because minimalism is cool. And fewer components = less styling pain.
Technically, it turned out well. There were almost no problems. Except a couple of things:
-
I didn’t know how to split components. I made a separate component for each page. Even though many could’ve been combined. I’ve since rewritten it properly — clean and efficient. But back then it was painful. I just made things harder for myself.
-
It's hard to plug in libraries when the project is a mess.
I wanted to make stars in the background of the welcome page. First I did it with ASCII symbols. Then a friend told me it was “too minimalist,” and I wanted to make the site more interactive anyway. So I decided to plug in React Bits.
And everything broke. Because the project wasn’t structured well.
If I had organized things properly from the start — I would’ve just imported a component and moved on. In the end, I got it working through hacks, but at what cost (actually no cost).
But I managed. And the site turned out exactly how I pictured it in my head.
It was an awesome feeling. When thoughts take shape — it gives you dopamine and strength to keep going.
👉 My site — ivan.space
It feels like a part of me. A glimpse into my inner world. At least, that’s how it seems to me. ฅ^•ﻌ•^ฅ
Conclusion
I thought I could fit this in one post — but no. So here’s the gist:
Web programming still feels like magic. But it’s magic anyone can learn.
These days building a site is like building LEGO. You come up with the instructions yourself. You swap parts on the go. Everything’s in your hands.
I truly believe everyone should try web dev. Especially if you're even slightly connected to IT.
It helps you understand what’s going on “in the kitchen” and teaches you to build working products (the right way from the start). And that’s a skill worth having.
-
Start with the basics — and go straight into practice.
Learn the tags, CSS, hooks, components — the fundamentals. You can do it in a day. Then just start making things. When a problem comes up — Google it or read the docs. -
A neural network is not your freelancer.
Don’t dump everything on it. Understand what it writes.
Otherwise, you’ll get lost in your own project.
Remember, after vibe-coding always comes vibe-debugging 🌚 -
Reading someone else’s code > writing your own.
When I needed to implement component overlays, I looked into the React Bits source code. Found tons of useful solutions and libraries (including Tailwind CSS).
Reading other people’s code, especially good code — is one of the best ways to learn.
You probably knew all this already, but it doesn’t really hit until you start building something yourself.
At some point you catch yourself missing something obvious — and that moment teaches you more than any tutorial.
That’s it for now. In the next post I’ll talk about technical details and how I ended up creating this blog — as the final stage of my journey.
Mostly, this post exists to motivate you — yes, you — to go out there and start learning web dev. Because it’s not just useful — it’s also fun.
So for now — take some time, dive into web, and build something of your own.
I’m sure you won’t regret it.
(And I’ll go have some tea. I’m tired of typing all this (´ж｀ς) )