By Mathéo Delbarre, 2nd year CS student at EPITECH Nancy, France
I've been publishing about ZamSync, a sync engine I built for district clinics in Bhutan that run on 2G and lose power mid-transfer. The response genuinely surprised me. Strangers from countries I've never visited opened pull requests. Someone took 30 seconds to write a message just to say the problem mattered. A developer forked the repo to build offline-first sync for password managers.
None of that was because I'm a great marketer. I'm a 19-year-old student with no following and no idea what I'm doing on that front. It happened because the problem was real.
That made me curious about something: what else is out there?
Not just projects. Problems too. The ones that don't have a solution yet, the ones where you've looked at the existing tools and thought "none of this actually fits." I want to hear about those just as much.
You don't need an idea to participate
This is the part I want to be clear about.
If you have a finished project, great. Share it.
If you have a half-finished prototype, share it.
If you have nothing built at all, but you work in a field where something is genuinely broken and you've spent years watching people work around it with spreadsheets, paper, or nothing at all... that is exactly what I want to read.
Describe the problem. Be specific. What breaks, when, for who, and why the existing tools don't fix it. Drop it in the comments or write a full article about it. That kind of description is not "less useful" than a GitHub link. It might be more useful, because someone reading it might have the technical background to actually build something.
The best open source projects I know of started as someone describing a problem clearly enough that another person thought "I know how to fix that."
What I mean by "real problem"
Not trying to gatekeep here, but there is a difference between:
- "I want a better to-do app"
- "District nurses in my region track patient visits on paper because the official system requires internet, and records get lost every monsoon season"
Both are real problems! But only the second one is what this thread is about.
The kind of problems I'm looking for share a few traits:
- The people affected don't have the resources to buy a commercial solution
- The existing tools are the wrong size, the wrong cost, or built for a different context
- The constraint is physical: no internet, old hardware, no power, a language nobody built software for, a population nobody considers a "market"
Some directions to get you thinking:
- Healthcare in low-bandwidth or low-resource environments
- Education where connectivity isn't guaranteed
- Agriculture, livestock, field data collection without cloud access
- Accessibility tools for disabilities or languages that the big platforms ignore
- Logistics in places where the infrastructure is unreliable
- Local governance, civic tools, community record-keeping
If you've seen something like this up close, write about it. You don't need to have the solution. The problem alone is enough to start a conversation. 💡
What to share
If you have a project: write about it. Any language, any stage, any size. What problem does it solve, who does it solve it for, what constraints did you hit, what would you do differently.
If you have an idea or a sketch: write about it. Explain the problem, explain your thinking so far, explain where you're stuck. Maybe someone in the comments has already solved part of it.
If you have a problem and no solution: write about it. Be specific. What is broken, for who, and why. That is the most valuable starting point.
Then tag your article #realworld and drop a link in the comments here. Or just describe the problem directly in the comments. I'll read every one, and I think other people here will too. The goal is to discuss, debate, push back, suggest approaches, show off work finished or not started. Not to do AI slop ;)
My contribution to this thread
I'll start. ZamSync is a Rust sync engine for district clinics in Bhutan: WAL replication with Hybrid Logical Clocks, Version Vectors, mTLS, and ChaCha20-Poly1305 encryption at rest. It runs in under 10 MB of RAM on a Raspberry Pi. It survives mid-transfer power cuts. It syncs exactly the missing delta and nothing more, because 2G bandwidth is not something you waste.
The full story is here:
The problem that started it was simple: district clinics needed reliable synchronization in places where connectivity and power are unreliable. ZamSync is one attempt at solving part of that problem.
Why I'm doing this
I've spent months inside one problem. I know that problem well. But I have no idea what equivalent things other people are quietly working on, in other fields, in other countries.
I think there is a whole category of work happening at the margins of the industry that never shows up in the feeds. Not because it's not interesting, but because it's not being built by funded startups, it doesn't have a product launch, and the people doing it are too focused on the actual work to write about it. Or they assume nobody else would care.
I think people would care, and I more think this community specifically would care.
So let's find out. What are you building? What problem are you watching that nobody is fixing? What have you tried that didn't work and why?
Drop it below. 👇
-- Mathéo Delbarre

