Ask HN: What's your app idea that you don't have the time or motivation to build

An actual implementation of Memex. That is, a web browser that records everything to a local cache, so that at any point later, you can annotate it, and give full and complete copies of what you've seen to your friends.

After a set period of time, things without annotations could roll off into the trash bin, waiting to be purged.

Given the size of modern disk systems, it should be no problem to store everything someone watches, reads, listens to, and decides to keep, for as long as they like.

I am a shader developer, often struggling with the incompatibility of underlying rendering tech (OpenGL, Vulkan, Metal, etc.). I built my own live-coding visual synthesizer using Kotlin script and OpenGL. If I had some time, I would port it to Google filament, so it is truly multiplatform.

Something along the lines of LinkedIn and the company announcement part of twitter.

I, and im sure many others, can do without the stream of BS that comes out of LinkedIn. But as far as an online job history and to search for companies, it's quite useful.

Seen today "I would dig an app that showed you how many lines of code you committed each day across every Github repo. Is it out there?" https://bsky.app/profile/ianlandsman.com/post/3lilcgtt7nc2c

I have a PhD in linguistics and have always been baffled by the lack of quality software for the types of data annotation workflows that linguists who do elicitation work with native speakers need to do. There's FLEx [0], but it's windows only and really not very good.

I'd love to build a modern, cross-platform replacement for FLEx. I just need to actually do it.

[0]: https://software.sil.org/fieldworks/

An org-mode datetree that toggles between chronological and reverse-chronological order.

A sane web archiver (SaaS or web apps do not count as sane).

Basically a docker image that can be hosted in the cloud (or on public facing machine), that allows you to "connect" any device, including your phone, to any other device. Kinda like ngrok but at the application layer. The idea is that you have a library that integrates with your web service running on localhost that defines an API, which then is publically accessible (with authentication) on the cloud container.

The use case would be just being able to interface with your stuff at home without a 3d party solution, in a programmatic way. So you can offload LLM inference, access stuff on your local network, and so on.

Additionally, the docker image should include ways to mesh with other instances, so you can form at hoc networks of devices.

I wish there was a way to open an MS Office file on my Linux Mint laptop with the real Word or Excel—no emulators that barely work, or office365 and no sluggish Windows VMs. Just a smooth, native experience.

geocities style site with AI. A website that gives everyone a blank page. Then the user can interact with an LLM/AI to build out the page. User1 might add a feed of X/Y/Z related tweets/bsky posts. User2 might have a section on their page for motivational quotes, a todo section and top news in the UK. User3 might have a different rock song everyday etc.

I thought of this earlier today and wasn't going to make it but now that I've typed it out it is tempting.

pages.ai is available (... for €95,840!)

I think it would be profitable to write a smartphone app that takes the floor plan of a trade show or exhibition hall with the exhibitors' booths shown, allows the user to select the booths he or she wants to visit, and solves the travelling salesman problem. I wonder if it could be licensed to trade show organizers for big bucks by offering kickbacks on all the extras. The extras could include bribes from exhibitors for the heuristic to favor paths that go by them, or that get users to visit them ahead of their competitors, or that push promotions to users redeemable at the booth but only within the next five minutes, or that lets exhibitors buy a list of the names of everyone who selected their booth, or lets them buy a list of the other booths those people also selected, etc..

Jira Viewer - connect to your self hosted Jira database, and view the content, navigate projects, search issues. Open source, modern, scalable, and free.

I still want a better todo list.

It's daily. Probably pulls stuff from a backlog. Maybe with deadlines, item 1 by 10 AM, (2) by 10.30 AM, (3) by 2 PM. Alternatively, it rolls a dice and tells me what to do next so there's no paralysis analysis.

Probably some task grouping, like if you're going to the bank, there's these tasks like buying a capacitor from the store next to the bank.

It'll break down a complex task into estimates. This can be done pretty easily with AI and RAG, and I'd be happy to pay small amounts for this.

A dating app but only for average or below looking people. Maybe it’s as simple as you get kicked off if you get too many dates? Or people could report you as too attractive?

A jobs app but candidates can only apply to one job per week. Employers must respond to any application before they’re allowed to get the next one.

Interests matcher- paste subscribed subreddits into form and get matched up with people with most common interests

P2P collaborative document/code editor. Not reliant on a central server.