alt Hacker NewsAsk HN: What Are You Working On? (April 2026)
What are you working on? Any new ideas that you're thinking about?
Comments
I'm building Tela (https://github.com/paulmooreparks/tela), a self-hosted relay that tunnels TCP services through encrypted WireGuard connections. The key difference from Tailscale and similar tools is that it requires no TUN adapter, no root access, and no admin privileges on either end. It runs entirely in userspace.
My initial motivation was wanting to RDP and SSH into my home workstation from a locked-down corporate laptop when I travel. I couldn't install Tailscale on the laptop, and I didn't want to pay for a cloud VM just to do SSH port forwarding. Now I use it to tie together half a dozen machines, both locally and on Hetzner & Linode. I can SSH and RDP into remote machines, host a git repo on one machine and access it from the others, and (optionally) share files across all of them on a local mount.
You run a hub (telahubd), register machines with a lightweight agent (telad), and connect from anywhere with the client (tela). All three are single Go binaries with no external dependencies. The hub never sees your traffic. It just relays opaque WireGuard ciphertext.
All binaries run on Windows, Linux, and macOS. There is also a desktop GUI app, TelaVisor, that wraps the client and enables remote management of hubs and agents.
It's Apache 2.0-license and pre-1.0 release, but I'm polishing it for a stable 1.0 release in the next month or so.
I'm also working on an enterprise-grade management portal that works with Tela, https://awansaya.net/
Grit Garden: https://grit.garden (https://github.com/pravj/wordle-garden)
Recently shipped this personal art project that turns daily Wordle attempts into gritty / struggle-filled stories, kinda similar to the emotional arc of the Wordle game play.
You can upload your own Wordle game screenshot to generate one for yourself.
In addition to completing what was once in the idea list, I got to learn about
- Prompt fine-tuning: Models are sharp enough to complete Wordle games quicker than human average scores, so I had to dump that down and get the average down.
- Karpathy’s Autoresearch: Experimented with auto-research for prompt fine-tuning, in addition to manual prompts.
- Vision models: While leading labs have multimodal models with quality visual reasoning, the benchmarks are still quite different for a simple Wordle analysis (reading what letters were yellow/gray/green); I also noticed labs/companies with separate vision models but their APIs lag significantly compared to what’s possible in developer experience.
- Video generation: For the last few days, I have been experimenting with automated video generation for the project's social handles. I'm still struggling with the right hooks that reduce the skip rates, but it's fun.
---
Additionally, working on an Apple Watch app similar to my Mac app on the same lines, [Plug That In](https://plugthat.in), i.e., notify before the device goes too low on battery, but with a twist.
I'm trying to get back to verifying some of my old fun ideas. I want to finally build my 3D QR cube (https://deriese.net/qrcubes.html?s=hn) by sending a design to a laser shop, and I also want to find someone with a few termocouples to verify my results to the coffee cup cooling problem (https://deriese.net/coffee.html?s=hn). If anyone wants to help, feel free to send me a message.
My mother is living alone in her house and we are getting to the point where she might not be able to live alone. I built "Still Kicking", a picture frame that monitors her motion and sends back basic reports and can detect falls and sleep quality to a phone app, to help give her more time at home.
It's just an mmWave sensor connected to an ESP32. But it works nicely, and I'm thinking of starting a company making them, though I'm not clear if the elderly would be ok with this minimal (no camera) intrusion.
It would just work out of the box.. the real one would have a small cell modem so it wouldn't need any networking setup, and it would act as a gateway if you have more than one in a house. There are industrial versions of this for nursing homes. This would be a bit more warm and fuzzy for home use.
I finished my immobiliser for old cars from the last thread. It fills a niche in the market for a no-wireless, no fob solution that will work on older vehicles without much CAN bus integration.
I have been brushing up some drawing skills for concept art, and exploring more embedded automotive product ideas for this niche of cars.
I got addicted to scrolling content on my phone, so I built a digital pet whose growth and well-being depends on you staying off your phone! This way, if I spend all night scrolling the browser, my pet will get depression.
Unlike similar apps such as Focus Friend or Forest, which use active timers to police screen time, my app is an inversion that works like an idle game; All screen time is tracked all day, (with double the punishments at night), and upon check-in, you get feedback on your device usage.
I've been writing a 'book' (more of an extended blog post that I'd like to put out for free) attempting to explain quantum computing to a layman-ish audience.
I sort of got inspired to do this after seeing so many QC PR posts on HN, and finding the educational material in this space to be either too academic, too narrow in scope, or totally facile. I think given the incredible hype (and potential promise) of this industry, there should be on-ramps for technically minded people to get an understanding of what's going on. I don't think you should need to be a quantum physicist to be able to follow the field (I am only an electrical engineer).
My book tries to cover the computational theory, the actual hardware implementations, and the potential applications of quantum computers. More than that, I want to be unbiased and stray away from what I feel is misleading hype. It's been a work in progress for about 6 months now, with a lot of time spent gaining fluency in the field. But the end is in sight! :)
Bare bones IM program inspired by bitmessage(.org): https://github.com/iljah/p2pIM Uses proof of work for spam protection. Consists of server and command line client written in python.
I'm working on Watch.ly - a remote human-in-the-loop networking and FS sandbox for AI agents like openclaw: https://watch.ly/
Also this week launching https://dirtforever.net/ which is an open alternative to RaceNet Clubs for Dirt Rally 2, since EA is shutting that down.
I'm also expanding the SDK and plugin space for https://fastcomments.com and am planning on adding AI agents because everyone expects that now :) a big challenge is building it in a way that doesn't make half the users mad. So I'm planning on marking any comments left by AI mods with a "bot" tag, and having the system email users on why it made certain decisions, with an option to contest that loops in a real person. I'm hoping this provides value to site owners while not angering real people. The agents could also just do non-invasive things like notify certain moderators when comments violate community standards in some way, or give out awards. I'm also hoping at some point I can run my own hardware for the LLMs so I don't have to share people's data with third parties.
I'm continuing to hack on Tiled Words, my daily word puzzle!
After winning the Playlin Player's Choice award I've noticed an uptick in players as well as some people sharing videos on YouTube which has been fun. I've got a few thousand people playing every day.
I just launched user accounts today so user's can now track their progress across devices and share their stats with each other. This ended up being a bigger chunk of work than I expected but I'm really pleased with how it turned out. (Though I launched it 15 minutes ago so I'm holding my breath for bug reports)
I'm fine-tuning my internal puzzle-building now with the goal of letting people use them to make and share their own puzzles soon!
A Minecraft Mod that outputs the Stats of the players onto a website (I made it use APIs, so an app or CLI tool would be easy to create later on).
This is my first Minecraft Mod and the first project i made that interacts with the network, has logins/accounts and does APIs.
I am really not good and thus the UI (far worse on mobile) and especially the Code is bad, but i would have never expected to get it working at all, much less this functional. But i am still far from done, i still want to improve the overall code quality, add the inventory and ender chest, achievements (maybe even custom ones so vanilla clients can earn and view them without having to change anything locally. IDK yet) and more.
If someone wants a small demo, i have it running on my server to test while I am developing at: https://grisu-ftp.de (If you find any issues lmk :)
While this is by far not as cool as the other stuff on here i still would like to show it off and gather some first feedback. This is my first Java project that goes above the Standard stuff in school like scanners/calculators and so I have probably done obvious beginner mistakes.
Trying to figure out how to get a job in this market for someone with sub 3 YoE in the industry :/ It's hard out there for juniors, y'all. I'm working at a company that I thought I could stay for years in, but my CTO left and now I'm shafted with basically all of their responsibilities - I'm not overly perturbed by this, as it's well within my ability, but I would much rather spend the next few years as an IC and really develop my skills as a SWE rather than jumping to manager this early... Also just getting an interview is insanely hard nowadays for some reason!
I recently rebuilt my homelab after moving countries, and in the process updated Proxmox to v9.1.6. Been playing with centralizing my databases into their own LXCs rather than creating an individual one for each application.
When I started doing this, I also decided to try Proxmox's new OCI compatibility, which seems to be working well so far, so I am removing all my Docker VMs and recreating the containers directly on my hypervisor.
https://larm.dev, an uptime monitoring service with a focus on reliability and reduction in false positives. I’ve been building it for myself really but I figure it’s worth sharing it with people in case someone else finds it useful too.
It’s also a lot of fun to work on. Phoenix LiveView dashboard, go probes running on 4 continents, connected to the backend using websocket tunnels. Clickhouse for reporting. Even did a CLI and an MCP for fun.
You can take the probes for a spin with the free response time checking tool and see how fast your site is https://larm.dev/tools/response-time
Finally being able to build something I've wanted for years: an alarm app that wakes me up 1 min earlier each day until I reach my wake time goal. Not sure if there is already anything like this out there, but I don't really care. I've been learning a ton about native Android development and so far it's been a painless process to be able to wake up earlier.
After going to a bunch of bachelor parties over the past few years, and planning a few myself - I built an AI-powered bachelor/bachelorette planning tool to help make it a lot easier then it actually is
Check it out at https://bartyai.com
I'm working on a fully offline, client-side train journey planner for UK rail - https://railraptor.com
When booking flights, I use sites like Kiwi and Skyscanner that let you do flexible searches - multiple destinations, custom connections, creative routes, etc. But rail search feels oddly constrained. All the UK train operators offer basically the same experience, and surface the exact same routes. I always suspected there were better or just different options that weren’t being shown. Where is the "Skyscanner for trains"?
After digging through the national rail data feeds, I decided to have a go at building my own route planner that runs completely offline in the browser. This gave me the freedom to implement more complex filters, search to/from multiple stations, and do it without a persistent network connection.
Now I'm finding routes that aren't offered by the standard train operators, connecting at different stations, and finding it's often easier to travel to different stations (some I'd never heard of) that get me closer and faster to where I actually want to go!
It's still a little rough and I'd like to add more features such as fares, VSTP data, and direct-links to book tickets, but wanted to share early and get some initial feedback before investing more time into it. So, thanks in advance - let me know what you think.
I've been working on https://game-pick.eu , a website for friends to easily decide on games to play together. It is voting-based and can show who has which game in their library to see who would yet have to buy it. I'm planning on adding features for finding new friends to play with also. It's my first real web project, so I'm excited how it will go.
Earlier today? My partner and I felled a couple of trees and bucked them into firewood to clear a spot on drier ground for our chicken coop, which had sunk halfway to China because we unknowingly landed it in a soup bowl three years ago when we moved in the winter when the ground was frozen. Also set and leveled four piers in the new spot for it to sit on.
Then slid it a few hundred feet across the lawn on composite deck boards we salvaged when we took a balcony down last year and landed it atop the new piers.
Then put the electric fence back up to keep the bears out.
Presently? A beer.
Still working on:
- https://github.com/rumca-js/Internet-Places-Database - Internet meta database
- https://github.com/rumca-js/Internet-feeds - list of Internet feeds
- https://github.com/rumca-js/crawler-buddy - crawling framework
- https://github.com/rumca-js/RSS-Link-Database-2026 - link meta from year 2026
- https://github.com/rumca-js/RSS-Link-Database-2025 - link meta from year 2025
Nitor - a discord clone with a shared/federated Identity layer, but self-hosted "servers/guilds". Trust model is to trust the guild server (e.g., so private channels work as one would expect with moderation capabilities), but to enable E2E DMs and friend/presence systems via guild servers as relays. Rust, Iced, Iroh.
Glyphcraft - a Minecraft mod (imagine if Thaumcraft, Ars Nouveau, and Hex Casting were smashed together)
B2B SaaS to host 3D scans of DataCenters and industrial plants.
Basically a google streetview tour of your Datacenter or large industrial plant.
You can do some nice things like draw 3D linework to trace the paths of pipes, conduits, eg : https://youtu.be/t8nRhWUl-vA add notes with markdown and html links at useful places in the 3D space.
We have add-ons for generating an 'xray' view floorplan to make it nicer to navigate a large space.
I think we are the first to have a web uploader that can preview and import .e57 panoramas, directly in the web page [ and skip the points if you dont need them ]
Currently in use by a telco in the Americas.
I'm writing an essay about how I use an ancient text editor, GNU Emacs, along with gptel, Gemini, some local models, yt-dlp, and patreon-dl to help me me study an ancient language, Latin.
I want to show how I liberate poorly aligned, pixelated PDF image scans of century-old Latin textbooks from the Internet Archive and transform them into glorious Org mode documents while preserving important typographic details, nicely formatted tables, and some semantic document metadata. I also want to demonstrate how I use a high-performance XML database engine to quickly perform Latin-to-English lookups against an XML-TEI formatted edition of the 19th century Lewis & Short dictionary, and using a RESTXQ endpoint and some XQuery code to dynamically reformat the entries into Org-mode for display in a pop-up buffer.
I intend demonstrate how I built a transcription pipeline in Emacs Lisp using tools such as yt-dlp and patreon-dl to grab Latin-language audio content from the Internet, transcode the audio with ffmpeg, do Voice Activity Detection and chunking in Python with Silero, load the chunks into Gemini's context window, and send it off for transcription and macronization, gather forced-alignment data using local a local wav2vec2-latin model, and finally add word-level linguistic analysis (POS, morphology, lemmas) using a local Stanza model trained on the Classical corpus.
This all gets saved to an an XML file which is loaded into BaseX along with some metadata. I'll then demonstrate some Emacs Lisp code which pulls it into an Org-mode based transcription buffer and minor-mode for reading and study, where I can play audio of any given Latin word, sentence, or paragraph, thanks to the forced-alignment and linguistic analysis data being stored in hidden text properties when the data was fetched from the database.
Lastly, I'd like to explore how to leverage these tools to automatically create flash cards with audio cues in Org mode using the anki-editor Emacs minor mode for sentence mining.
A Ruby-inspired typed programming language called Sapphire: https://github.com/sapphire-project/sapphire
I was reading the fantastic Crafting Interpreters book, and been wondering what it would be like to design a language from scratch. I really enjoy using Sorbet with Ruby, so wanted to design a small language with Ruby's object model, and a gradual type system.
Despite not knowing much programming language theory, I was able to make a surprising amount of progress over a couple of weekends using Claude Code, including building a simple version manager for the language - https://github.com/sapphire-project/facet
Working on tooling to help make working with agents in parallel easier, with minimal tools/no deps - https://github.com/andrewhathaway/ag.sh I don’t want to manually manage worktrees, tmux sessions, branches but want to remain in the terminal.
Also recently built a home energy cost/consumption display for the TRMNL - https://andrewhathaway.net/blog/ambient-cost-display-for-oct...
https://folio.photo -- it's a photography portfolio that doubles as client gallery delivery, with white labelling, password protection, and other features you don't get with the average file sharing platform.
I'm hosting my personal gallery with it: https://captures.moe
I am building devlens.io, an open source tool for codebase visualization tool for easy onboarding and easy PR review. The most interesting and loved feature of the tool is blast radius i.e., If I change this component, how far will the effect be propogated ?
github repo if you wanna check : https://github.com/devlensio/devlensOSS website : https://devlens.io
I'm making web chat using captive wifi portal on esp32.
I have been making GTK applications so that people can manage MergerFS and LUKS encrypted hard drives without knowing how to be a sysadmin.
The use case is kind of neat. RAID is great and can mostly solve these problems, but people don't have SATA hardware that can handle the workload well, plus they aren't ready to manage an array like that, and they don't like having to use specific sized drives, etc. Another major issue with those setups is you need to be careful because an IO error that you don't recover from will be very difficult or impossible to recover from because of the layers of LUKS combined with LVM.
With MergerFS you just use regular file systems that are separate, but they get combined into a single mount point. That means each disk can just be a different LUKS encrypted drive and if you need to recover data it's isolated to that one disk and much more manageable. You can also take any disk and plug it into another machine as needed and grow or shrink the storage pool as needed.
MergerFS has options and settings to help you determine how files are spread across the drives, such as least space used or which disk has the most of that directory path already.
My app (Chimera) automates the unlocking of the disks, mounting and some data migration if you want to remove a disk from the pool. I plan to add some rclone features to help provide easier backup options to places like Backblaze, AWS, or a remote server in general.
So far so good and I was surprised at how well Opus had been handling Gtk and pkexec.
Let me know if you guys are interested I am close to pushing some RPMs and DEBs, in addition to the standard Python stuff.
Im building https://trypixie.com to legally employ my 7 year old child, save on taxes and contribute to her Roth IRA.
I also built https://statphone.com - One emergency number that rings your whole family and breaks through DND.
Since blacksmith labs and phoenix ai - the ai assisted automotive r&d, security and compliance engineering tool has been on the backburner for a while, I've been helping a couple of blockchain projects and built some tools, besides tinkering with my real-time EVM contract and transaction heuristics and classification engine.
- ETH Watchtower: a real-time EVM monitoring tool with heuristics and classification of contracts and transactions: https://ethwatchtower.xyz
- P2P SSL VPN provider/consumer tools using a blockchain as announcement and settlement layer: https://github.com/rnts08/blockchain-vpn
- OrdexNetwork: https://ordexnetwork.org, I've built https://ordexswap.online and https://ordexswap.online/wallet/ as well as an Umbrel variant of a self-hosted wallet.
- Waya Wolf Coin v3: Helped the team to compile binaries for linux, and modernizing the libraries: https://github.com/rnts08/WWC3-Linux-binaries / https://github.com/Waya-Wolf/WWC3
- Low Cap Exchange algorithmic trading bots with machine learning and automatic ghost trading, because I wanted to see what the most common shapes are on smaller exchanges: https://github.com/rnts08/low-cap-exchange-trading-bot
However, I am really looking for Sr. DevOps/Platform Eng/SRE/System/Network Admin/Infra Engineering or similar, full-time or contract work, see https://timhbergstrom.pro for contact details.
The same thing I do every night - try to make FreeBSD work better on EC2!
Ok in all seriousness, right now I'm tracking down an issue with the ENA network interface which results in sporadic packet loss. Triggering the issue is hard and seems to require a large number of TCP segments being pushed to the NIC very fast. So far I've found that my reproducer stops reproducing when I turn off write combining on the MMIO space used for low latency queueing, which is... just a little bit weird.
Im working on a all-in-one event management platform for studios, event planner etc..,
It consists of CRM, Expense tracking, Equipment Management, Event Gallery( photo share, Face Detection based download, Guest Upload) etc..,
Currently working on moving it from cloud supabase to self hosted version.
I've been building Jikimi, a privacy-first parental control platform. It started because my kids were spending too much time on the computer, only stopping if we asked them to. If we forgot, they'd just keep playing. We were also worried about online predatory behaviour like grooming, bullying, that kind of thing.
So I built an on-device OCR engine (PaddleOCR) that reads screen text locally and feeds it into an AI sentiment analysis pipeline. No screenshots leave the machine. We now get alerts if there's detection of concerning interactions. The client is written in Rust, with DNS filtering, game detection (Steam/Wine/Proton), and screen time enforcement built in.
It started as a home project that worked really well. My wife suggested other families wouldbenefit, so I've been building it out as a product. The client shipped on Linux first, we're a Linux gaming family, with Windows coming soon.
There are many more features I haven't touched on. Would love feedback from other parents who've dealt with this space. The goal is to protect children and empower parents with tooling that's transparent and effective.
Working on Gaming Couch, a web-based local multiplayer party game platform. It's like a lovechild of Jackbox Games and Mario Party: https://gamingcouch.com. Three months ago, back in December, Gaming Couch hit the front page of Hacker News (https://news.ycombinator.com/item?id=46344573). We've had an amazing time since, with each month more and more people finding the platform, enjoying the games and giving awesome feedback!
At the moment working on the 3rd party development tools so in the future anyone can make their game dev dreams a reality and make a simple and fun multiplayer party game for the Gaming Couch platform, ideally in only one weekend!
If you're an interested game dev that would like to beta test the dev tools, hit me up either here, via Discord (link available from https://gamingcouch.com) or by emailing me at gc[dot]community[at]gamingcouch[dot]com!
The TL;DR of Gaming Couch:
- Currently in free Early Access with 18 competitive mini-games.
- Players use their mobile phones as controllers (you can use game pads as well!)
- Everything is completely web-based, no downloads or installs are necessary to play
- All games support up to 8 players at a time and are action based, with quick ~one minute rounds to keep a good pace. This means there are no language based trivia or asynchronous games!
Synchronous P2P file sharing tool with post-quantum encryption and virtual mount point (https://keibisoft.com/tools/keibidrop.html)
Both peers mount a virtual FUSE folder. Files shared by one side appear in the other's folder in real time. You can open, copy, and browse your peer's files as if they were local. Files go directly between devices over encrypted gRPC. (by default it tries over LAN, then direct IPV6, then uses a data relay).
The hardest part has been making git repos work through the FUSE mount between peers.
(Been developing the tool for 12 months now, very close to a full release)
Space Trader!
Imagine mixing Magic: The Gathering, StarCraft and Civilization’s hex grid combat.
There’s multiplayer but I haven’t put the server anywhere yet.
Check out the introduction here:
https://github.com/williamcotton/space-trader/blob/main/docs...
Clone the repo:
npm install
npm run dev
I'm working on an app called Limberly. It focuses on health and ergonomics for sedentary workers - probably most of us here :)
It is scientifically proven[1] that sitting is detrimental to our health, with increased mortality rates. The primary way to reduce the negative effects of sedentary work is to move.
This means doing sessions of resistance training (gym), running, biking, but also taking micro-breaks during work sessions and performing light exercises and stretches.
Research has shown[2] that taking short breaks during work reduces fatigue, and in some cases actually boosts performance.
There are plenty of running and gym apps out there, so Limberly focuses on the last part - helping you take micro-breaks, reminding you to change your posture between sitting and standing, changing which hand holds the mouse (if you're into that) etc.
It is still in early development, so if you'd like to help test and shape the app as we go, please sign up for the waitlist and I'll add you to the testers group. Feel free to also DM me here with any questions or feedback.
Oh, I am also writing a series of articles that explains it more in depth: https://prodzen.dev/articles/building-limberly-part-1-we-re-...
1: https://pmc.ncbi.nlm.nih.gov/articles/PMC10799265/
2: https://journals.plos.org/plosone/article?id=10.1371/journal...
I believe that AI-powered software development means we need to fundamentally rethink how we preserve code quality.
Model output volumes mean that code review only as a final check before merge is way too late, and far too burdensome. Using AI to review AI-generated code is a band-aid, but not a cure.
That's why I built Caliper (http://getcaliper.dev). It's a system that institutes multiple layers of code quality checks throughout the dev cycle. The lightest-weight checks get executed after every agent turn, and then increasingly more complex checks get run pre-commit and pre-merge.
Early users love it, and the data demonstrates the need - 40% of agent turns produce code that violates a project's own conventions (as defined in CLAUDE.md). Caliper catches those violations immediately and gets the model to make corrections before small issues become costly to unwind.
Still very early, and all feedback is welcome! http://getcaliper.dev
I'm making capability security for distributed systems. The primitives and engine are both open source. Primitives: https://github.com/Hessra-Labs/hessra-tokens Engine: https://github.com/Hessra-Labs/hessra-cap
It's built using biscuits and written in rust. I'm really into it. Using capability security as a model makes building things feel like they snap together a lot more naturally. At least for me.
I've also got a blog post describing it in more detail: https://www.hessra.net/blog/what-problem-led-me-to-capabilit...
I struggled finding mock exams when I was studying for the Einburgerungstest in Germany so i built a super simple site to offer folks some mock exams to practice with https://www.einburgerungstestpractice.com
Working on m experimentall alternative to PHP's composer. Some projects took 2-3 mins to update with composer, only to fail at the end. Wanted to rework this, but the composer code is big and I honestly didn't feel like touching more PHP than I had to.
So I built my own package manager that's almost ready for alpha.
https://crit.md - a CLI tool for reviewing AI coding agent output like a GitHub PR.
I got frustrated with Claude Code and Cursor producing plausible-but-wrong changes with no easy way to annotate and push back, without making a full PR. crit makes the review stage fun again!
Works on both plans as well as code itself. It’s been very rewarding hearing from folks who use it, everyone has been very kind! My most successful side project already :)
I'm back to searching for numbers that are palindromes both in decimal and in binary. [0]
I had an insight the other day, that as I fix the n least (and most, it's a palindrome!) significant decimal digits, I also fix the remainder from division in 5^n. Let's call it R. Since I also fix by that point a bunch of least (and most) significant bits, I can subtract how much they contribute mod 5^n from R, to get the remainder from division in 5^n of the still unknown bit. The thing is, maybe it's not possible to get this specific remainder with the unknown bits, because they're too few.
So, I can prepare in advance a table of size 5^n (for one or more ns) which tells me how many bits from the middle of the palindrome I need, to get a remainder of <index mod 5^n>.
Then when I get to the aforementioned situation, all I need to do is to compare the number in the table to number of unknown bits. If the number in the table is bigger, I can prune the entire subtree.
From a little bit of testing, this seems to work, and it seems to complement my current lookup tables and not prune the same branches. It won't make a huge difference, but every little bit helps.
The important thing, though, is that I'm just happy there are still algorithmic improvements! For a long while I've been only doing engineering improvements such as more efficient tables and porting to CUDA, but since the problem is exponential, real breakthroughs have to come from a better algorithm, and I almost gave up on finding one.
[0] https://ashdnazg.github.io/articles/22/Finding-Really-Big-Pa...
I'm working on Ruly, a daily number/logic puzzle where you set rules on a grid.
My goal is to make a simple yet interesting procedural and replayable puzzle. It has a couple of weekly variations: on Saturdays you need to break a rule to score max points, and on Mondays there's an added memory aspect which brings variety to the game.
It's mostly vibe-coded which lets me focus on game design and testing. The next step is better onboarding/tutorial and more intuitive UI.
Worked on some features at open reader, a local-first PDF TTS reader that highlights the words spoken and uses the excellent local kokoro tts engine.
Got fed up with web tech, it's so slow and clunky, so made my own version in python and qt. I changed the design to be based on a doclayout llm, so you can skip or include things like tables and references easily.
It now works so beautifully fast, it's code is readable and simple, no apis or multiple services. Just a qt app, some local llms that can run on a decent cpu and word-leven highlighting and playback selection.
https://github.com/thepycoder/projectwhy-tts
I can listen to papers now!
I'm working on https://concludia.org, a site that helps groups of people collaborate on arguments and conclusions. I don't really have any revenue plans for it currently as I suspect it will be rather niche -- I certainly wouldn't mind if it tops out as a small community of users -- but I've found it super useful in various contexts at work and at home.
You can read more about it over at the site, but it allows you to construct and validate arguments in a graphical form, and it has truth/proof propagation so you can see whether a conclusion is currently considered valid or contested. You can create counterpoints where you think the argument breaks down, and strengthen arguments from there. Some upcoming plans are to allow users to validate arguments for themselves, like mark which parts they understand and agree with so they can collapse that part of the graph, and to add more mcp capability so that LLM can help you construct and validate new arguments.
I’m working on my own markdown IDE / Google Docs competitor with an AI agentic editor that coaches on strategy in addition to proof reading. I made it as a side project. Designed it entirely by taking screenshots, annotating them, and giving feedback to codex. Basically applied everything I learned utilizing Claude code a d codex extensively at work to this side project to see how fast I could ship something that felt complete. Check it out: https://clarus.page
I’m working on an ATS system that integrates LLMs for helping the recruiter in writing job descriptions, parsing CVs to Markdown, making summaries of CVs and suggest which ones match the best the job offers giving pluses and minus: https://beehive-ats.com/
I'm working on a sandbox for Linux terminal work with UX similar to Python virtualenv: https://github.com/wrr/drop