What do you guys think about htmx? It's looking like one of the most sensible ways to move forward with making dynamic applications, to be honest.

What I like in particular is that there is no prescription for "other things" tied to it. It literally is just Hypermedia On Whatever You'd Like. On top of that it has no complex requirements for how to deal with types on either side of the equation; the client side is a browser that can read hypermedia and the server can produce hypermedia.

this is giving me alpine.js vibes.

i can totally see a move to this - i think it's why things like Svelte are becoming popular since people want more brain-dead simple front-end code without having to learn a completely new API.

where did you hear about this?

Indeed, I think alpine.js fills sort of the same niche, where things that require entire stacks are reduced to something... less.

I heard about it maybe 6 or so months ago randomly when I was googling around while working on some Haskell web stuff. One of the things that came up (I think in a reddit post?) was htmx.

I got reminded by it again a while ago when someone highlighted it on HackerNews so I decided to dive in a bit more and see what I could find out about it. There are some good presentations by Carson Gross about it.

Lately I've also been on a simplification kick overall because I'm just so tired of everything in terms of complexity. We don't use React and your usual suspects for work (but rather PureScript compiled to Erlang and a library for writing LiveView applications in PureScript) but it's overall way too complex for what we are doing and so I've been trying to get my simplicity fix elsewhere.

Edit: The complexity really is everywhere; I'm getting tired of functional programming, higher-level languages and basically everything that is non-obvious in terms of what is going to happen memory-wise and in terms of roughly what the processor is going to do. As an industry I think we're spending a lot of time on things that aren't even really buying us much execution speed, development speed, control or otherwise.

This looks really good. Markup languages are one of my favourite to use because they are so straightforward. Getting rid of the scripting side of the markup language is probably a good move, but a tough one, because they would need to be able to keep it simple. Otherwise the markup will become incomprehensible to write.

It's serendipidous I found this thred. I have had friends in the webdev scene chatting about htmx for the past few years trying to get me interested and I finally decided to take a look last week.

I have become a big fan of Inertia.js with Vue3 components, it being building JavaScript apps the monolith way really spoke to my love of the traditional request -> response way of building websites. What I disliked about Inertia is that in order to have SSR you have to also have a node service running alongside your app that you pipeline responses through, this adds a point of failure, a magic black box and one more step between your code and your user agent. The hydration approach of modern JS frameworks also feels overly complex to me.

HTMX feels so much nicer. I can return to the simplicity of user agents making requests and just responding with server rendered HTML and use Alpine.js to fill in the gaps.

I haven't used it, but from what I have read I like HTMX. At my last company we had a powerful home-built system that let us put a lot of extra tags & markup in the PHP templates and the frontend would get the interactive parts that we need without diving into too much JS stuff. It was built when jQuery was popular and "outdated" by the time I joined the company. But it was very fast to get stuff done with. I think HTMX looks like a cool way to have that out there for anyone to integrate with. I'll be using it on my next side project.