It depends on the target product.

I'm working with JS for already 25 years. Tried all of the frameworks, and continue on doing it. And every time I try something new, the refactoring flow turns most of them into NextJS (if it's very UI rich or customer facing or something very web-oriented), or Vite+React+Tailwind (client) and Hono (backend) if it's more of a tinker toy needing more custom solutions.

The boilerplate with NextJS is cleanest (compared to all the other frameworks) and API is the most straightforward one, and you can safely ignore the vendor lock in. Its just a pretext to hate on NextJS. They all have some kind of a "vendor" lock in. Be it a vendor-or-a-specific-approach-or-whatever-lock-in.

And Vite+React+Hono — simplest to set up for quick experiments, and very powerful with minimal boilerplate. Will probably create a starter for this one, as I have been using this stack quite a lot lately.

EDIT:

You can pretend vanilla JS is all you need, but then your app grows, then you suddenly need types, and state, and more events and their handlers, and SSR or something else. Thus React has been the most stable bet for quite a while for me now.

> The boilerplate with NextJS is cleanest (compared to all the other frameworks) and API is the most straightforward one, and you can safely ignore the vendor lock in. Its just a pretext to hate on NextJS. They all have some kind of a "vendor" lock in. Be it a vendor-or-a-specific-approach-or-whatever-lock-in.

The vendor lock-in on NextJS is certainly much more egregious than other frameworks. They have their own undocumented build flag to give different build outputs that Vercel uses vs the build outputs that are documented. Hosting nextjs on your infrastructure is not as simple as sticking it into a docker file as with most frameworks

And I would also push back on the idea that every framework has vendor lock in. Remix was so focused on "using the platform" that it basically willed itself out of existence. It's no longer even a framework. Just part of the react-router library. I've also used Astro which is a framework of similar complexity and feature richness as Nextjs and certainly has no "lock in". At least as far as

I'm curious if you've tried Lit on the frontend, and if so, what you think about it.

I have tried it. And would like to reiterate – everyone should use what they like.

But for me Lit is too OOP. It feels like Angular. And that all in turn feels like Java. It's just so heavy and constrained (not saying it's a bad thing though). Too much boilerplate for me.

The whole paradigm is different and does not match my preferences. And while subjective, I do believe React with TS, Tailwind, zod, react-query and zustand is the best stack delivering the best balance of boilerplate and code-delivery and easy of use and entry level and dx.

Well this crashed my Safari like nothing else in quite a while!

Well, the landing page is obviously vibe coded (I assume with Bolt based on the colors).

Found out you can use WASD to navigate and Space to shoot.

I would've probably switched from JetBrains to Zed already. But Zed has no vertical tabs support.

I can't believe people are ok with horizontally layed out tabs.

Since you're here — I'll just pipe in.

Here in this article, the author, failing to comprehend the domain differences, is applying the same approach to call a function everywhere. Of course it won't work.

The fallacy of nextjs is attempting to blend function domains that are inherently different. Stop doing that and you will be fine. Documentation won't work, it will be just more confusing. Blending edge and ssr and node and client-side into one is a mess, and the attempt to achieve that only results in layers upon layers of redundant framework complexity.

Sounds like you wouldn’t be a fan of React Server Components in general then since blending domains is its whole point.

Blending domains is great. Blending domains where you don't get logging at some levels because your framework is incompetent is not.

Consider middleware.ts as a root middleware. Nothing is stopping you from creating your own chain (which is trivial) in there. I mean, that would eventually work the same if nextjs implemented that feature — there would be a root somewhere.

That doesn't answer parent's question.

People expect "middleware" to mean a certain thing and work a certain way.

  middleware = fn(req) → next(req).

express/koa give you the use() chain. next.js gives you one root, but nothing stops you from chaining yourself. same semantics, just manual wiring.

  type mw = (req: Request, next: () => Response) => Response;
  
  const logger: mw = (req, next) => {
  console.log(req.url);
  return next();

};

  const auth: mw = (req, next) => {
    if (!req.headers.get("x-auth")) return new   Response("forbidden", { status: 403 });
    return next();
  };
  
  function chain(mws: mw[]) {
    return (req: Request) =>
      mws.reduceRight((next, mw) => () => mw(req, next), () => new Response("ok"))();
  }
  
  export function middleware(req: Request) {
    return chain([logger, auth])(req);
  }

root is given, chain is trivial. that’s middleware.

Nothing trivial about that implementation in my mind - need to keep track of where middleware is registered, reduceRight is non obvious.

I expect these things to be standardized by the framework and all the sharp edges filed off - thats why I go to a framework in the first place.

The reduceRight is just a bit of cute FP code golf. All it’s saying is that chaining an empty list of middleware yields an ‘OK’ response, and that the first middleware is passed a function which, when called, executes the remaining middleware chain, and so on. It would be obvious enough if written out as a for loop, or via direct recursion.

(My username has never been more appropriate!)

This is terrible and not worthy of HN front.

Terrible from the front-end side of implementation: - performs worse than your average arbitrary-amount-of-rows-that-won't-fit-on-the-screen library (it should perform the same no matter if its 1k, 1m or 1mm rows) - is seemingly buggy - is pointless on its own, because THIS demo is a client-side demo, and no one loads that much data on the client-side.

Revisit this when this demo is performant AND data is loaded from the backend.

Ignoring that, every front-end JS developer should explore these kinds of libs and also try to implement them themselves, because they're basically front-end 101.

That is fine as long as your forms are simple text inputs and buttons. Now plug in drag-and-drop and multiple file uploads and selects and checkboxes and radios and more of these various inputs you're in the world of pain.

I've recently, once again, gave native inputs a chance in a new project. It lasted as long as I've described in the first sentence. And I've been in the frontend world for 20 years. Trust me, you don't want complicated native forms.

And react-hook-form is just what you need (albeit it also is boilerplate-ish, so I always end up wrapping it up in a simpler and smarter hook and component).

edit: Same, in a sense, for HTMX. It's ok for simple things. But eventually you may end up trying to build a house with a fork. The fork in itself is not a bad tool, sure. But you also don't need a concrete mixer with your morning toast.

Enter this in the console:

document.body.onwheel = (e) => e.stopPropagation();