I counted last week: 6 projects sitting on my GitHub, and not one of them is finished.
I started most of them after 10 pm in my room in Dhaka, full of energy, usually minutes after watching someone build the same thing in a 12-minute video.
Here lies each one: why it died, in one honest line, and the one real thing it taught me before it did.
The weather app that never saw rain
Why it died: I spent three evenings choosing colours for the layout and never finished reading the weather data.
The app could show you a beautiful empty box. Inside that box was supposed to be a temperature, pulled from a free weather service. Getting it meant reading the service's documentation and understanding the shape of what came back: labelled values with names like "temp" and "feels_like".
What it taught me: what an API actually is, in my hands instead of in a definition. A program asking another program a question, and getting a structured answer back. The weather app died so my other programs could ask questions properly.
The chat bot that only ever talked to me
Why it died: it worked on my laptop, and putting it anywhere real scared me off.
On my screen, it was charming: type "hello", get a reply; type "fees", get the semester fee dates I had fed it. It answered maybe 8 questions, all of them asked by me, in my own wording, which is the easiest exam a bot can sit.
What it taught me: how a conversation is just a loop with memory. Take in a line, look for what it matches, answer, ask again. Under the charm, that is all it ever was. Later, when I met the same loop shape in class problems, it felt like meeting an old neighbour.
The Snake clone that ate its own tail
Why it died: one collision bug. When the snake turned fast, it judged itself dead a moment too early, and I could not find why for a week.
The game ran. The snake moved, ate, grew. Then, on quick turns, game over, for no visible reason. I read my code until the lines stopped looking like mine.
What it taught me: how to debug by watching numbers instead of guessing. I printed the snake's head position every step and compared it with where my code thought the wall was. The two disagreed by exactly one square. Off by one. It was my first real off-by-one, and I have recognised its face in other bugs ever since.
The blog with exactly one post
Why it died: I spent all my energy choosing how it would be built and had nothing left for the writing.
The single post is titled "Hello world", and it is two sentences long, one of which is "Welcome to my blog." Building the blog taught me a stack of new words; writing the second post required the one skill I was avoiding, which was writing.
What it taught me: a blog post is just a text file with a title, and a page is that file wearing a design. Once I saw my own post turned into a page by a few lines of code, websites stopped looking like magic and started looking like plumbing.
The price tracker that tracked nothing
Why it died: the shop's website changed its layout, my program's map of the page stopped matching, and every price quietly froze.
For about a month it checked the price of a book I wanted. Then the numbers stopped changing, and I believed them for an embarrassing while before checking by hand.
What it taught me: other people's websites owe you nothing. If you build on a page you do not control, you are renting, and the landlord can repaint without telling you. It also taught me to make programs complain loudly when data looks dead, instead of smiling and repeating it.
The flashcard app I needed most
Why it died: my exams arrived, and I stopped building the study tool in order to study.
The irony is not lost on me. This is the one I still feel. It showed a question, waited, showed the answer, and asked if I had known it. Simple, useful, aimed directly at my weakest subjects.
What it taught me: how repetition can be scheduled. Known cards come back rarely, shaky ones return tomorrow, and the whole system is just dates and honest self-grading. I now revise with exactly that method, on paper, because the app never grew up. The idea survived its own project.
The part where I am supposed to redeem myself
There is no redemption section, because there is no redemption. I finished none of the six. If I started a seventh tonight, the honest odds say it joins the others.
Here is the catch I did not expect: several of these lessons only arrived months later, when a class problem or a new project suddenly rhymed with an old corpse. At the time, each project felt like pure waste. They were not waste. They were tuition, paid in evenings.
Finished would have been better. I am not going to dress it up. But the graveyard was not free of value just because it was free of launches.
Your count
So, honesty for honesty. Open your profile, count the ones you never finished, the ones with big readmes and two commits and silence.
How many unfinished projects are sitting in your graveyard? Give me the number.
P.S. Everything I write stays free — that part doesn't change. But if this one was worth something to you, you can buy me a coffee ☕ (one-time or monthly, both work). It covers the API credits and hosting behind these experiments — and honestly, it's also the nudge that gets the next one started on the evenings I'd otherwise skip it. — Salam
Top comments (0)