Been building a small native app on the side and keep wondering if it's worth finishing.
Feels like AI is slowly eating everything — and the assumption seems to be that the future is all cloud, all connected. But then you also have local LLMs taking off, which makes me think maybe it goes the other way.
Anyone still building local/desktop tools? Or should we just stop and move on?
As AI starts handling more and more of the implementation work, I feel like developers are being forced to lower their standards around fully understanding and controlling the codebase.
A few years ago I cared deeply about architecture consistency, naming, structure, reviewing every line, etc.
Now the amount of generated code and the speed of iteration is becoming too large to realistically track in detail.
At some point, are we slowly giving up programming itself and shifting into product direction + validation instead?
Can experienced developers here actually accept that transition?
https://reddit.com/link/1tensvb/video/xipcs52zmg1h1/player
It's Friday! Sharing some results from last week — curious to hear what you think.
I built this tool mainly for my own use, not really planned as a product at the beginning.
My workflow involves a lot of image handling:
I found myself constantly switching between tools — some online (easy but requires uploading), some CLI (powerful but not convenient for quick tasks).
So I decided to build a small offline tool to simplify this.
It currently supports:
It's built as a Windows desktop app (C# / WPF), and everything runs locally.
Still pretty early and definitely not perfect, so I'm curious:
Any feedback is appreciated.
(If needed, I can share the link in the comments.)
https://reddit.com/link/1t5xfnt/video/38ejqlolkmzg1/player
I kept running into situations where I needed to share PDFs (resumes, contracts, etc.), and realized that just blacking out text doesn’t actually remove the underlying data.
So I made a simple tool that scans PDFs for sensitive content and lets you review and permanently remove it before sharing.
Works fully offline, supports batch processing, and gives you control over what gets removed vs kept.
Would be interested to hear if this is something others run into, or how you’re handling it now.
I’ve been working on a desktop tool for a while and finally got it to a point where it’s usable.
Here’s the weird part: when people actually land on it, conversion isn’t terrible. But getting anyone to discover it in the first place feels almost impossible.
I don’t have an audience, no ad budget, and I’m not comfortable spamming links in communities.
So far I’ve tried:
Result: almost no visibility.
It’s starting to feel like the real problem isn’t the product, but distribution — and I don’t really know what the next step should be.
For people who’ve built niche tools before:
Where did your first real users actually come from?
Was it search, communities, direct outreach… or something else entirely?