Live data from Hacker News

Ask HN: What are you working on? (May 2026)

news.ycombinator.com

221–230 of 1001 posts

Re: Ask HN: What are you working on? (May 2026)

#221
I've been vibecoding* a calendar-based personal note-taking thingy: https://github.com/cu/doneski

The idea is that each morning, you click the "New Day" button, and your Todo list along with other notes carry forward from the previous day to the new one. When you accomplish something, you add it to the Done section. Other sections can be added as needed. I have been using a text editor and/or shell script for this purpose for about a decade, but have been inspired to make it into an app now that I can delegate the boring bits of app development. It is not quite done yet, but it's getting close to being usable.

(* To the inevitable downvoters, this is in part an experiment to get familiar with what SOTA LLMs can handle. With the intent of comparing it to local LLMs once I get my Strix Halo set up as a coding assistant. I only code as a hobby currently, and have too many other hobbies, and this app wouldn't exist without something else doing the heavy lifting. That said, this is a pretty low-stakes application and I don't commit any code that I haven't reviewed and don't understand.)

Re: Ask HN: What are you working on? (May 2026)

#222
I vibe-produced a website for an (earnest?) pastiche of Final Fantasy games (i.e., for a game that does not, and probably never will, exist). I noticed that Nano Banana could generate reasonable facsimiles of Square Enix's promotional render style and ran with it. Next up: faking some gameplay or a few shots from an FMV-style cutscene.

https://findfantasyxviii.com

Re: Ask HN: What are you working on? (May 2026)

#224
https://surmai.app/

I'm working on a Personal / Family travel organizer. Started as tool to allow me and SO to plan a trip together. There's been steady progress over the last couple years. Focus on privacy and ability to self-host. Of course, there is a managed version if one doesn't mind me having access to their data.

Re: Ask HN: What are you working on? (May 2026)

#225
I'm working on https://fdeploy.com

fDeploy is a self-hosted Windows deployment automation tool — a lightweight, on-prem alternative to Octopus Deploy. It consists of a Server (Windows service with a Web UI) that orchestrates releases, and Agents installed on target windows machines that execute deployment steps (IIS sites, file copies, scripts, etc.) across environments.

Re: Ask HN: What are you working on? (May 2026)

#226
I've been working on RVW, my adaptation of the standard transformer model that is capable of online continual learning without catastrophic forgetting. I finally published the first pre-print of my early experiments: https://doi.org/10.5281/zenodo.20064617

Now I'm working on expanding the work into more parameters and improving performance. I just finished an extremely harsh test of a Nemotron-flavored RVW that consisted of stretches of a random assortment of domains interspersed with long runs of single domains. Across all of it the model didn't forget (and actually improved on some of the more challenging domains). PPL on SmolTalk is still in the ~18 range, which I'd like to get lower, but this is all with only 4B params.

Currently, I'm training a Llama 3.2-flavored RVW with only about 2B params to see how that turns out. Depending on results of that, I may take it to Gemma 4 next.

Re: Ask HN: What are you working on? (May 2026)

#227
I recently switched to developing VST audio plugins and I'm loving it. Already done 3 [0][1][2] and I want to keep doing this if I manage to generate some income from it. I develop them in Typescript and then convert them to C++ with Webview, this way I have a web demo of the plugin that is almost identical to the one you get for the DAW.

[0]: https://technokick.com/ (Techno Kick synth)

[1]: https://riviera-demo.surge.sh/ (Reverb effect)

[2]: https://ya3.surge.sh/ (TB-303 synth clone)

Re: Ask HN: What are you working on? (May 2026)

#228
I've made https://rankr.click

It's a little web application that allows for the ranking of all kinds of abstract entities. Think of the merging of Goodreads for books, Vivino for wine, Letterboxd for film, etc. This will allow you to instead rank whatever you want across a variety of different categories in a single place.

Using your rankings across all these different fields, you can draw analysis of what you like, and in future I'd like to add a little personal (not an ad) recommendation engine to help you find new stuff based on your actual interests across loads of different categories.

From a technical point of view, its been a great learning opportunity on how to fully host a complete stack using an opiniated, but cross-platform orchestrator, allowing me to host this wherever (bare metal VPS, homebrew system, cloud provider) in a flash.

Re: Ask HN: What are you working on? (May 2026)

#229
I just started on an open source and open weight supervised learning model to recognize japanese kanji characters drawn on the screen.

I have a working prototype written in Julia which is a very simple neural network. The input is in vector format so traditional convolutional neural networks don’t work out of the box but I swapped the convolution layer with a path simplification algorithm and it worked extremely well. Like 20 samples per character (from a set of only 5 hiragana during prototype phase) was enough to get 100% accuracy in a test collection of 5 samples per character after only 30 iterations of training.

I plan an working with free and open data, which I don‘t think exists for japanese kanji characters (at least not in vector format; KanjiVG only has one sample per character and I need dozens) so I also build a crowdsourcing web site to collect data from random people on the internet.

I am planning to run some more experiments with my prototype model before I release the crowdsourcing web page to an actual server though.

Model prototype: https://github.com/runarberg/kantoku-prototype

Crowdsource app: https://github.com/runarberg/kantoku-collector

Re: Ask HN: What are you working on? (May 2026)

#230
We're working on Webhound - budget controlled long-running deep research. You set a budget and Webhound will use that much in compute/LLM tokens to research your prompt, with built in verification cycles and optional added verification budget. Every claim is cited with evidence and a direct link to the tool calls that produced the claim

The goal is to build a deep research product for actual researchers, since we believe that it is an extremely powerful product that is still nascent but has enormous potential - which we've already seen with some early users.

https://webhound.ai

Post reply on HN