pseudo_meta

Treesitter is fantastic. It has builtin support in nvim, and there are a lot of plugins that make use of it.

My favorite is nvim-treesitter-textobjects which gives you dozens of new targets for vim motions, such as a function call or the condition of a loop.

I use Brave, and for me it's really just the least bad option.

Firefox-based browsers do not support macOS automation (AppleScript/JXA). Safari lacks features/extensions. Orion/Vivaldi had bugs any time I tried them.

From the Chrimium-based browsers I tried, Brave blocks ads, supports PWAs, the crypto stuff can be turned off, and is stable. Brave does not excite me, but it's good enough.

Plain and simple: It's an AI-enhanced "I'm feeling lucky".

As someone who uses the "I'm feeling lucky" regularly, this is definitely an improvement.

Money does not make you happy, but it can solve problems.

Don't aim to be rich, but try to have enough.

Love rss, but the upside of not having an algorithm determine your content consumption quickly results in a fire hose of content.

Sadly, filtering features seem to be only available for paid subscriptions of online services, or for self-hosted solutions. Or are there solutions I am not aware of?

- native browser bookmarks: things I visit regularly and want to access quickly, like the website of my bank. - read-later-links: I send it to my todo app, and read it the next day or delete it. No piling off links I never read. - topic-specific links like "tool to extract tables from PDFs" go into my note taking system (in my case Obsidian), where I can also write notes on the links properly

Oh thx, must have missed that. Guess at least that saves me some time to switch to the newer API in the future.

API is noticeably slower for me, sometimes up to 10x slower.

Upon some digging, it seems that part of the slowdown is due to the gpt-5 models by default doing some reasoning (reasoning effort "medium"), even for the nano or mini model. Setting the reasoning effort to "minimal" improves the speed a lot.

However, to be able to set the reasoning effort you have to switch to the new Response API, which wasn't a lot of work, but more than just changing a URL.

I've only used reflog really to look up a hash and reset back to it. Is there anything else about it one should learn?

"Waiting for the AI agent" is the new "waiting for code to compile," it seems.

Is it possible to trigger the expander differently?

I already have zsh-magic-dashboard running on empty enter.

Landline phone. Never used it, the only thing I got there was spam, marketing or survey calls. Twitter for reasons that probably do not need explaining.

If low-tech counts: Contacts. They are just so much hassle compared to glasses, the imagined better looks are not worth it.

Read-later services. Just another pile of things to manage. If I really need to read something, I add it to my todo app.

A formatter with little to no configuration, similar to black.

A linter what the reasoning for all rules is well explained, similar to shellcheck or eslint.

Both ideally integrated in an LSP, which also has all the common features of a modern LSP.

> Show all images on the page in a grid with size and megapixel tags. Useful for manual scraping. Would you mind sharing that one?