A new OpenAI LLM model came out : gpt-5.1-codex-mini ; it's 5x less expensive than gpt-5 or gpt-5-codex. Interesting!
It's minimal, but I'm posting things.
AI context should look like a top-down funnel, become more and more precise and relevant to each step of the task at hand.
A new OpenAI LLM model came out : gpt-5.1-codex-mini ; it's 5x less expensive than gpt-5 or gpt-5-codex. Interesting!
I'm going to try Neovim; I've avoided this as much as I could given how much time I poured in my vimrc config. I hope it goes well.
Firefox has an incredible feature which is more convenient than browser profiles : tab containers. You can store difference user sessions in them.
After an afternoon of work, I noticed my mouse has been turned off the entire time; it's so nice to work mostly using the keyboard :)
That LSP logging file which was slowing down my Vim so much, it's has been turned on for over a year...
My Vim was feeling more and more sluggish. After profiling I realized lsp_logging was enabled. Thousands of IO operations to a log file was bogging everything down. Disable LSP logging!
It's a little disheartening Java development is never a nice experience in Vim.
I have no doubt we are in an AI bubble, so pick open-source solutions ; when the ecosystem will break down, at least those projects will remain.
Just set up HomeAssistant and it was super easy on a Raspberry Pi 3 Model B.
You can program your AI agents with OpenCode in the same way you can program your text editor with Vim : you feel in control, and it's fun :)