I find the idea of building (also with) frameworks these days annoying and boring because of churn. The problem with building a framework is that you are not going to be building it for general use, only your use cases. Once you want to expand the scope of your framework because you had a moderately good experience using it in your last project, you'll find yourself wanting to expand things but in a way that breaks the old, so now you have churn.
It's funny because when you build with front end frameworks like React and Svelte, you'll find that things have changed because they want to expand what they can do therefore they also cause churn. I find it hard too believe they couldn't find better ways to do things than to restructure how people define components or hook them up.
NodeJS which is used to build front-end websites and apps also had that churn problem where they decided to move to esm, lol I noped out of it and dropped NodeJS completely. This is just busy work that i'm not interested in.
You know what hasn't changed fundamentally to cause churn? HTML, CSS, Javascript. You could take a website built with jQuery and it'll still run in people's browsers today. htmx seems like the best we are going to get and definitely deserves to be the best, I don't think it is a library that can fundamentally change given it uses attributes on html. 95% of websites can get away with htmx and web components and eschew React, Svelte, Vue and all the other frameworks.
You don't even need NodeJS for htmx, you just drop it in and it works, anyway I digress this isn't a advertisement for htmx.
Churn is a big problem with front-end development. I went from server side rendering (php) to client side rendering (frameworks) back to server side with some client side rendering where needed. If I want to create a webapp, Leptos is something I can leverage, it uses wasm, so it isn't exactly something to use like a hammer. If all I need is servers side rendering, I use Snapfire to leverage tera templates this is zero churn. Deploy and never care again, it will work as long as html, css, js don't fundamentally change.
One of the things I see people talk about when it comes to libraries like htmx is that it is for solo or small teams. I believe one would only say this as a cursory thing. They don't actually believe that, it is a in the moment thing because it sounds right. This is not unlike an LLM, so yeah, the kind of person who would say this is a bot. Teams will suffer if they adopt practices of a solo dev, derrrrr. It doesn't matter what you use. If you haven't figured out how to componentize, I would regard you as a newbie. Small, Medium and Large teams all benefit from components.
Avoiding frameworks like the plague should be the norm. If you want to build an SPA, sure go pick React or Svelte or whatever, that is the right use case for those. Even then I'd still be hard pressed to use them. I would still be wary of frameworks since they have been known to change whenever they feel they need to expand. There's no good reason for me to accept the churn just because the library wants to change to bring new things in. If HTML did that shit, all of these mf framwork developers would be complaining.
Occasionally I will chat with ChatGPT about what it would look like if we had a proper library for zero churn frontend and the road leads to libraries like htmx. If I wanted to create my own library today, it would look like htmx. It handles a wide variety of cases and has a really small runtime with no compilation nor build time needed.