Hacker News Reader

Comment by sailorganymede | Hacker News Reader

I am a big fan of Vite. But I have zero clue what those other tools are. I swear to God, I just put my head down to do some work and all the sudden, frontend tooling has evolved. I wonder if there is a push towards a "boring but works" stack.

View on HN

> But I have zero clue what those other tools are.

The incorporated tools are actually really amazing:

- vitest, an ultra fast test runner. After using a lot of others, including jest and node's built in one, I love vitest.

- oxlint, replaces eslint but is compatible with its file format and ultra fast, since it isn't written JavaScript. I tried biome, but I found oxlint to have more rules and the eslint compatibility was nice.

- oxfmt, replaces prettier and is faster since it isn't written in JavaScript.

- rolldown, replaces rollup and is compatible with it but it is much faster since it isn't written in JavaScript.

The above are my go-tos in new projects anyhow since they are killer good and fast.

I wish Oxfmt supported plugins. Prettier's plugin API is one of the worst APIs I've ever worked with. I'm eager to switch to a different formatter with a better plugin API (and I need plugins).

Have you tried contributing a system for that?

I was excited about Oxfmt until I tried it and found that it's mostly intended to be a Prettier replacement. Ugh.

I'm curious to know what you were hoping for. Are there features you're looking for in a formatter that prettier/oxfmt don't offer?

I was hoping for something more flexible. I suppose the whole point of Prettier is that it's not flexible but I find 90% of the times the code is actually uglier when using it.

Try Biome then. But Oxfmt is still faster.

being worked on : https://github.com/oxc-project/oxc/discussions/21936

Rolldown is used in Vite (non plus) though.

I don't get how a test runner can be "ultra fast". Surely all the time is taken by the tests, not calling the test functions?

At work we've tried switching to vitest, and it's 1.5-2x slower than Jest (I think it's related to our very large and circular module graph), so performance is very much a your-mileage-may-vary thing.

The initial selling point was performance, but then they gradually realized that a lot of the slow cruft in Jest was necessary for correctness, and now it's about the same performance as Jest (obviously may vary in some specific situation).

However vitest is still great! Selling points now are stuff like:

- shares config with vite

- works with ESM out of the box (I think Jest still doesn't)

- integrated browser testing mode that is very nice

- overall just has a ton of nicely integrated features

I do recommend turning off "isolate" for as much of your code base as possible when it makes sense. And I recommend ensuring "maxWorkers" is being used properly, I prefer something like 60% of my totals cores as the number of workers to use. And use a top level vitest start so it properly runs all the packages in a pipeline rather than as separate vitest runs (which would mess up the maxWorkers optimization anyhow.)

I have my 3000 test project suite completing in 15 seconds on my MacBook Air M3. It is pretty sweet with that setup.

> Surely all the time is taken by the tests, not calling the test functions?

Calling tests has overhead. Also knowing how to schedule and parallelize tests with dependencies is not as "simple".

Say this again when you have worked with Jest, one of the worst and slowest pieces of software I've ever worked with.

And Jest was itself a huge step up from what came before (Jasmine, Mocha...)

I switched to Jest from Mocha and my memory is that Mocha was much faster.

You'd be surprised how slow the JS ecosystem can be.

You'd be surprised https://github.com/artemave/node-test-runners-performance-co...

I expected Bun to be among the fastest and yes, I was not surprised.

> Vite, Vitest, Rolldown, tsdown, Oxlint, and Oxfmt

You already use the first, the second you should be using and has existed for years - it's a jest-compatible runner integrated to Vite.

Rolldown is a part of Vite already (since version 8, alongside other implementing tools like lightnincss).

The ox* suite are native replacements for eslint and prettier. The latter has existed for years. Both of them aim to be drop-in compatible soon.

This is a boring stack.

FWIR, prior to the purchase by Vercel (or was it Cloud flare?), Vite+ was going to be a convenient, no fuss, low movement way set up projects that they sold to businesses for funding.

> This is a boring stack.

I think the boring stack would skip the need for most of these. That is 5 different tools to write TypeScript in order to get some HTML, CSS and JavaScript for things that are not covered by modern web standards. Not including the frameworks itself which mostly have the same goal.

That's a pretty disingenuous take on the purpose of TS and misses the point of having a stack in general. The tools actually cover a lot more than that.

Yes, if your goal is to get HTML/CSS/JS in any form to a web client then you don't need this stuff.

If your goal is to ship a web app that stands a chance of surviving multiple devs, production traffic, etc then you need more tools. Testing behavior (vitest), enforcing code standards (oxfmt and oxlint), optimizing network calls (e.g. tree shaking from Rollup), etc. all go beyond what you describe.

This stack is as boring as it gets without literally using the tools it consolidates and replaces or dropping a requirement that necessitates them.

Cloudflare.

the other tools are for testing, bundling, linting and formatting. Previously you would use different tools from very different open source projects for these things, with different configurations, update cycles etc. Now it's all covered by one simple toolchain. Vite+ is basically the "boring but works" stack, while also being more performant and with less configuration required.

This is the latest emerging "boring but works" stack.

"Latest emerging boring but works" sounds like an oxymoron.

Or sarcasm.

So something can only be "boring but works" if it was created before today?

Yes, it's not boring if it's the new hotness

Okay, then how can something become boring? It needs to start as the new hotness? Does that imply that the people who use non-boring software are the ones who determine which things get to graduate to being boring?

More or less, yeah?

Yeah, boring implies that it is predictable, proven and mostly unchanging. In software those almost always come with age.

Unfortunately, yes. It's a backward looking metric, like how economists can only call a recession/depression after it's been X months of bad metrics, even if everyone living in the first month of one feels the economy going poorly. We'll only know if these tools are "boring but work" at some point in the future.

The hope is always that "this time will be different". Maybe these tools are so awesome that they'll become a de facto standard and, therefore, "boring". It's JavaScript, so I'm not holding my breath, though.

That's what I liked from Bun's proposition. A single binary that just works. Hopefully the others take notice.

Honestly, I hate that Bun is vibe-coded and seems amateurish in many ways, but it’s still an excellent tool.

so instead of looking at the result, you are complaining about (a minor aspect of) the process.

Jarred is an excellent engineer. He'll deliver.

I’ve bounced off Vite because I don’t see a big advantage over esbuild. (I also use Deno.) What do you find useful?

It's a linter, a code formatter, a tester, and a bundler. What exists in your "boring" stack that's more boring than that?

> I wonder if there is a push towards a "boring but works" stack.

Yea, this.

eslint -> oxlint (but in rust, and fast)

prettier -> oxfmt (but in rust, and fast)

webpack -> vite (...I have thoughts; but vite is popular enough so w/e)

rolldown -> tsdown (ts support, but in rast and fast)

jest -> vitest (works with vite)

It's basically taking the last decade of established conventions and supporting ts, rewriting in rust for speed, and making it all interoperable.

A "push" towards a boring stack? That won't happen because the hype cycle trends towards new tools like water down a river. But if you're looking, I can't recommend Rails enough in 2026. Built on web standards, it's quietly pushing the framework forward and is so much less maintenance than modern JS apps.

I'm currently experimenting with a stack combining assistance, ASP.NET Razor Pages, and htmx, specifically to avoid dealing with the complexities of the modern frontend stack.

AI eliminates a lot of the tedium, grinding tasks, while HTML partials and htmx complement each other perfectly.

Not having to manage frontend framework, SSR states is incredibly liberating. Sprinkle in some vanillla js and redraw specific parts of the page as needed... done

Are you open to opportunities?

not at the moment :)

You're on a good path.

While consulting, I have seen .NET + htmx in production, but it was an internal tool used by thousands of folks (HR system).

The 1ms response times and quick browser rendering due to no JS bloat put a smile on my face.

I have since been planning to give C# backend + htmx frontend a go for a CRUD demo.

I use oxfmt and oxlint, the first JS linter I've used. For a long time, I avoided Eslint due to how strict it was, but even though oxlint uses eslint, I've been ok using it. I haven't used vitest before though.

And I think Webpack should be discontinued by now, because Vite has replaced it in every way. Maybe to get other projects to switch, they should make a Webpack-to-Vite migration script.

I guess I'm just an old man pumping my fist and yelling at the clouds at this point, but I think compartmentalizing every possible aspect of development and then making each one of those as complex as possible lead AI in the bureaucratic deterioration of the trade.

Fear not, engineer! Recently some smart folks invented something called BackRub that’ll soon solve all your problems! Soon you’ll be able to find information in even less than time than it takes you to type a comment and out yourself as clueless!

Reply on news.ycombinator.com