I too got laid off at the end of April. So what I've been working on since then has been AI. But of course, being the contrarian that I am, I'm trying to make small language model AI usable. The end goal is to complete all the pieces I need to create an AI empowered tool that you can point at an authorized Network, and it figures out on its own. What needs to be monitored. Learns what normal looks like on your network. Build your dashboards. Working both as systems monitoring and network monitoring. So right now my GitHub is full of both things that I need, for personal use, in this goal. Also with things that I need because they will become tools for the final product. https://github.com/linuxrebel. You'll get to see my contorted mind at work

Yesterday I reached a stopping point in a project I had been working on over several weekends. I wanted to connect my illumos server directly to my Japanese ISP through DS-Lite, and ended up writing and publishing two Rust crates along the way.

The tunnel itself worked, but stock illumos could not manage the DHCPv6 prefix delegation required by my connection. I also spent quite a bit of time tracing a four byte ABI mismatch between a 64 bit Rust process and a 32 bit illumos service.

There was no happy ending, but I learned more about the illumos network stack and distilled the experience into a blog post: https://skalski.dev/connecting-my-illumos-box-to-a-japanese-...

Still in beta with Verse Draft: https://versedraft.com - An all-in-one writing studio where fiction writers can keep all the details for their universes in one place while crafting stories, novels, movie scripts, TV series, or stage plays.

One surprising finding is how few people are interested in real-time voice interaction (transcribing, dictating, conversing with assistant), so that's been removed for now.

I've been building https://promptship.dev for a while now. Basically a cloud where the operators are the agents not the humans. The platform allows agents to attach DBs, clone envs with data (instant prod like env for testing, etc), per-env access tokens for agents, append-only logging/traces for agent actions.

It's been quite useful for our small team and we plan to open this to public beta, soon.

I'm working on a tool for visualizing biological pathways, pulling data from the Reactome knowledge graph for now. Trying out different graph visualization algorithms, it's an interesting problem space as they are all approximations to solving a problem that I believe is NP-hard-- arranging a graph such that all vertices are distinct and you have the global minimum of overlapping edges. Next step is implementing a novel algorithm that I found described in a dissertation that doesn't share its code. Unfortunately it's not deployed yet, but if anyone seems interested I will be sure to share it on HN when it's in an acceptable state.

Interesting to see the distribution of projects here. For people building something with the intention of eventually making it a business: what was the strongest evidence that the problem was worth pursuing before you started building? I'm especially interested in cases where the initial idea changed substantially after talking to users.

Im working on what i call a spreadsheet data platform. It looks like a spreadsheet but all the components are plugins which means the user can provide (even write their own).

- Dont like the limited functions available in formulas, write your own. Yes i know this isnt new, just highlighting a starting example.

- Dont like how values are formatted in calls, write your own.

- Dont like templates, write your own script that generates the "empty" or not so empty spreadsheet.

- Dont like clicking on icons, write your own scripts or plugins that execute in context, which makes snippets to do basic things become very brief, because theres no need to load and save the file with the spreadsheet or even select the range with a query.

- Want a terminal, i have one. Everything in a spreadsheet is a file, just like *nix. Read/write a cell, is as simple as reading /cell/A1.txt or /spreadsheet/123/cell/A1, because the environment has smarts like environment values (eg SPREADSHEET_ID) that are used.

I have many other ideas which are nearly there.

https://github.com/mP1

I’m working on PhotoFig: https://photofig.com, a recently launched product-photo editor for online sellers who need to clean up hundreds of images.

It was inspired by my wife having to carefully mask product photos for her online shop. Existing tools could remove backgrounds automatically, but didn’t give her enough control when the result needed correcting. So masking is a first-class feature, with SAM-assisted selection and a precise manual brush.

The backend and storage are self-hosted on hardware I own. Along the way I’ve had to solve some interesting infrastructure problems, including writing a Kubernetes operator that bin-packs GPU workloads across workers. I’m planning a technical write-up on the architecture.

I work on a suite of open source public transit software projects, collectively called OneBusAway (or OBA). The software is used by millions of transit riders every day around the world, including in New York City, the Seattle area, San Diego, and Washington, D.C.

Through the non-profit that collectively owns OBA, I've been working on a managed services offering to help transit agencies adopt the software, which has historically been beloved by users, but a real pain to set up and maintain. The software, cleverly named OBA Cloud, turns what used to require a month-long engagement with a transit consulting firm into a 15 minute, wizard-driven process.

The best part is that since all of the software is free and open source, a transit agency that decides they love OBA but wants to be locked back into an expensive contract can always ask a consulting firm to run the software for them.

Since last month[1], we've added a number of great new features to OBA Cloud, like broadcasting critical alerts to all riders in the system via push notification, and support for Live Activities on iOS.

We're hosting four new public transit systems over last month, and are continuing to make terrific strides on our other marquee projects:

The iOS app update went live, and our riders are generally loving it: https://apps.apple.com/us/app/onebusaway/id329380089

The new Android app maintainer is gearing up to release the biggest new version of the app in a decade: https://github.com/onebusaway/onebusaway-android/

Our GSoC interns are continuing to do amazing work on their projects, including the next generation OBA API server, Maglev: https://github.com/onebusaway/maglev

We're always looking for more folks to contribute to the project, across a variety of disciplines (ie not just engineers!): https://ossvolunteers.com/organizations/open-transit-softwar...

[1] https://news.ycombinator.com/item?id=48888489

I have been building a social network for Twitch, called Chattr. It is very Twitch specific as you log in with Twitch and it will automatically post when you go live. When your stream ends, it will automatically post the VOD to your followers feed. Also, if anyone who follows you on Twitch joins Chattr, then they will automatically follow you on the site too. This makes it much easier to move your community over.

You can also tip and receive tips for your posts, which can then trigger a stream alert.

You can use your own custom emotes in posts and your Twitch subscribers can also use them.

I'm really excited to be building this and we have 135 users signed up.

https://chattr.online is the link and the app is on the Play Store and Apple App Store