Live data from Hacker News

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

news.ycombinator.com

291–300 of 828 posts

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

#292
I made a database[1] a few months ago and now I'm beginning work on a Github replacement (for me). Here's what I'm thinking for the UI[2]. I'm not trying to have feature-parity out the gate, I just want something that doesn't give me the ick. I self-host public repos with cgit but that's too basic and there's no built-in support for private repos.

The existing Github alternatives...look like someone drawing Github from memory.

   [1]: https://disc.sh / https://disc.md
   [2]: https://social.coop/@netopwibby/117227113793136781

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

#294

About fifteen years ago a puzzle game called Trainyard came out for iOS and took my dev team by storm for a few weeks. It ceased getting updates some years ago and isn't on the app store anymore. So I came up with a variant on it and just put up https://www.geoffcanyon.com/lightyard.html I need more puzzles to include in the base pack, if you come up with something interesting, export it and send it to me!

trainyard is/was an amazing game.

Agreed!

Despite the fact that trainyard has been defunct for ~8 years, I didn't want to simply clone it.

I tried to give a similar, but not the same, experience -- I'd be very curious to know how close you think I came.

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

#295
I'm building notmutt (notmuch + mutt). A tabbed, async terminal mail client with unified inbox, with included html viewer(with image support), mcp server, lua scripting. I tried using neomutt with unified inbox and I ended up messing up my maildirs every so often. i also tried adding async tasks(for sending) to neomutt, but there was no interest from the project, which made me realize that there is a lot more to it than just sending async mails.

The goal is to have a very fast terminal mail client where you don't have to wait for any operations and that is tightly integrated with notmuch(a common mail full text search engine) tagging engine, and other workflows(like crm queues and ai summary's if enabled in the config).

There is a document that explains why it was made in go. It has the best notmutt cgo bindings, and there are other reference terminal clients I can use as reference. althogh himalaya(rust) also looks interesting.

I would love to get some help with it, but I don't know where I should post it to get people interested.

https://github.com/fishman/notmutt

https://fishman.github.io/notmutt/

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

#298

Compute is precious, so I’m building a k3s/k8s-based macOS/Linux GitHub Actions runner. I’m spending like $25/mo at least in GHA minutes. Then working on securing and tinkering with my homelab. And little utilities that make repeated deterministic functions easier for me and agents.

What steps are you taking to improve its security?

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

#299
Two things seriously, one in (partial) jest:

A copy utility that can efficiently deduplicate as it copies, even when the filesystem structure is different. Whilst there are many utilities that can deduplicate, they often either rely on files being on the same filesystem, or having the same structure. If a duplicate is detected, it can either be omitted or hard/soft links created to files in one or more reference locations. In practice I have found this to be invaluable for merging and efficiently consolidating multiple hard drives with partial backups of photos and music, maintaining the complete trees of all sources on a single destination, or only copying files that are new.

Secondly, an efficient protocol for text applications over a network. Think a combination of TN3270, SSH and JavaScript, with a very lightweight client/server architecture. Whilst SSH is great, latency kills many practical applications - this attempts to solve that. It also intends to make applications very simple to implement. As a proof of concept, I implemented a simple forum with login, optimistic posting, quoting, a text editor with mouse support in about 600 lines of code. The protocol, client and server themselves are agnostic to any particular UI toolkit.

Thirdly - microfish, poor person's microfiche using laser printers and autoencoders to preserve readability and machine recoverability of high density prints. Since consumer digital archival media seems to be dying...

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

#300
post #47

https://fabulae.orbilii.com - A Latin language audiobook containing both volumes of A.D. Godley's "The Fables of Orbilius" which are entertaining intermediate Latin stories written at the turn of the 20th century for use in British schools. The audio itself is in delivered with the restored classical Latin pronunciation and contains spoken nuances like vowel nasalization, elision, and prodelision. I used a wav2vec2 m…

Woah this is so interesting. Confirmed it myself, you actually implemented elision!!! I really like your idea of capturing provenance in-document as well. This is humans speech, right? If you can do forced alignment on Latin, it gets me wondering if one could use a wav2vec especially turned to convert IPA into speech somewhat deterministically to compensate for how there is not heaps of training data in Latin.

Haha, thank you for using multiple exclamation marks!!! I'm thrilled that someone is as excited as I am about this feature, it made all the time I spent manually correcting/marking all of Gemini's elision hallucinations worthwhile :)

The recordings were made by my friend and Latin tutor. I used an Italian wav2vec2 model fine-tuned on Latin[0], which works pretty well for Latin only content, but falls down with mixed English / Latin. I found that some of the slower, larger Meta models worked fine for this use case though.

Some folks on the Latin language Discord said they got pretty good results using Kokoro TTS and X-Voice for creating synthetic Latin, as you suggested using IPA.

[0]: https://huggingface.co/lsb/wav2vec2-base-it-latin

Post reply on HN