Given 6 months have passed since this article has been published, would you also pick MicroOS? I'm currently setting up a single-node k8s cluster on Hetzner but I went with Talos. Curious about your takes, IaC and k8s is new terrain for me.
I've started a side business to support local artisans.
I want to help people making a living from what they've created with their hands. A major focus to do so is to connect with local artisans and learning from their perspective.
On the technical side, I'm developing a static site[0] with Astro / React / Shadcn. The main interface is a OSM leaflet map with a custom calendar timeline.
As a software developer, this is my main entry in the world of agentic engineering. I'm using OpenCode with models from OpenRouter. One of the surprising insights is that there are often free quotas for new coding models (e.g. MiniMax). It's a very interesting time for building.
I notice a lot of the suggestions are news-oriented. However, the thing I like about HN are all the non-news gold nuggets about tech. I'd love to find such gold nuggets for other industries as well.
I think George is right about vibe coding. It's great for rapid prototyping, but in my daily programming on a production codebase, PRs must be reviewable, so I have to understand what I'm changing even more so.
I love the example with git blame for the current line. Helix is my daily driver alongside lazygit, however I much appreciate a tighter integration of Git in Helix. A colleague showed me what he can do with magit in emacs, and that was some next-level stuff (e.g. cycling the buffer through the Git history of the file). Previously I was using Neovim, but I'm really happy with my switch to Helix. So much less config, keybindings I find more intuitive and multi-cursor editing. For agentic coding I'm looking into Aider and OpenCode. I expect a tool like that to join my terminal setup with Helix and lazygit.
This is such a nice project utilizing the advances in TTS models. Just converted an ePub in German via OpenAI. This gives me way more flexibility in reading / listening to books.
I'm very curious if this works. I'd like to have this product, but so far it's hard to find independent assessments and (long term) side effects as well are unknown. Also I think it's unclear if a one time treatment is enough or if it has to be reapplied regularly (e.g. after a professional dental cleaning).
A use case could be that anyone can provide a piece of content with the accompanying embedding and this can be used for semantic search. I.e. the search engine does not have to compute the embedding of everything, just the query.