vecty VS hlive

Vecty

It depends on how specifically you don't want to write HTML/CSS/JS and how broad your definition of "frontend" is. There are a handful of all-go frontend frameworks such as Vecty and Vugu of varying maturity and completeness. Then there's other libraries that more or less have you write HTML tags in go, such as go-app.

It draws on Go's syscall/js library as inspiration, which is pretty powerful (I wrote a pretty popular React-like framework using it a while back.)

A note on wasm: I'm building a hobby project with it right now and have tried different frameworks, I tried vecty which is nice to compile but full of bugs and unexpected behavior. I'm now on vugu which works better but is still harder to work with than a JS framework.

https://github.com/hexops/vecty/ is a framework for developing dynamic web frontends in Go. It's not production-ready, and will likely get some rewrites once generics drop, but it's pretty neat (I'm a contributor).

There's Vecty, though it's likely to get a bit of a redesign once generics drops.

How does it compare to https://github.com/hexops/vecty or https://github.com/bep/gr ?

Reminds me of Vecty from the GopherJS project: https://github.com/hexops/vecty .

I do a lot of wired stuff with server-side virtual DOM, and I've found Go a good fit. I don't think inheritance is a problem with designing a component library. I've used struct and interface embedding to share common functionality. (see Tag and Tagger, then Component and Componenter as sort of examples).

Poking through the repo at https://github.com/SamHennessy/hlive now. It's particularly interesting to me because I wanted to write a virtual DOM for something that did serverside rendering and clientside via WASM. I wanted to be able to parse a template language much like jsx or Angular templates for a delcarative syntax. A friend of mine I worked with years ago and I set up a 3 day hackathon. We went through some design work, decided we didn't have time in our lives to take on a project like this, and settled on using Go's html templating. I tried to reuse a lot of the concepts using a two pass method (since you can't hook into the parsing but can always register DOM listeners) thinking I could accept a little inefficiency in exchange for a lot less work. Finally, the weekend was over. We had some interesting ideas collected with some innovative new thoughts on state management, but he dove into his next project using Vue instead, and I haven't had time to continue.

I'd appreciate it if you tried to build something using this: https://github.com/SamHennessy/hlive