> could we throw the definition into context and tell the LLM to avoid it?
Every IMPORTANT RULE WHICH MUST BE FOLLOWED STRICTLY to not to do something from my AGENTS.md and skills are violated without any hesitation from LLM. Glorified autocomplete doesn't work this way. Positive rules are more stable on this regard.
Without tailwind LLMs tend to use inline random colors, random sizes and radiuses. Project becomes an unmaintainable design soup really fast. Even with tailwind first priority is to define dark/light theme with semantic colors. After it's a smooth sail (sort of, routing, forms, state ownership and persistent url state are still an alien concept for clankers).
I'm quite versed in CSS but tailwind has convenient abstractions on top. It's like "why you need C if you have assembler". I don't remember when I wrote raw CSS last time. Even global styles I tend to express via @apply.
Hm it's interesting perspective. The paper adds sorting into the mix. Like sorting moves cells and thus prone to issues with references. For me sorting always moves values in cells and I even can't imagine how you could sort formulas. It doesn't make sense.
ArrayList is a very generic (pun not intended) structure and could be stretched quite freely in any direction with useful property of owning underlying slice. Like readonly preallocated ArrayList is a thing.
Since pipewire landed in arch all my issues with sound magically vanished. It's a really great piece of software. Maybe it's even worth to change the distro.
It's funny I'm currently creating a TUI. A mutt like mail client. I didn't have any second though to consider GUI for the project because it's so ridiculous.
It's inherently cross-platform (msys under windows). It allows to copy any part of UI. It doesn't require dependencies other than curses.
reply