Hacker News Reader

Comment by rk06 | Hacker News Reader

The Unified toolchain is an extremely ambitious project. when Rome announced their plan for unified toolchain, I expected it to fail as the next HN reader. and turned out to be right.

Bun is also attempting it. Thye have made tremendous progress but they are also competing against node, and thus I don't expect to for bun to go mainstream.

However, despite the difficulties, I strongly believe that Vite+ can achieve it.

I strongly recommend all readers here to watch Vite documentary[0] that got released less than a day ago for vite's history and bacground of Vite+

[0] https://www.youtube.com/watch?v=bmWQqAKLgT4

View on HN

Well, Rome failed because they run out of money while the project weren't finished yet. While Vite+ seems to have most of the things done already, so I'd consider it a success for now, what is left to see if that the companies using Vite already are willing to pay for Vite+.

> "Rome failed because they run out of money"

The sestertius isn't what it used to be.

Rome also failed because it was attempting to build everything (transpiler/bundler/linter/etc) from scratch.

It was also unfortunately timed. When they started they were competing against webpack, but right around the start of the project compilers written in more performant languages like ESBuld and SWC start to take off and out compete Rome before it even got off the ground.

TIming was hardly an issue since they didn't produce anything important. moreover, their first focus was on formatter, which can be as low priority for any dev as it gets.

Even Vite+ is focused on vite and rolldown first, and formatter last.

And that is on purpose!

Sure, but a large part of why they never managed to produce anything useful is because SWC/ESBuild basically made the whole Rome architecture obsolete and killed all momentum around the project before they got anywhere.

what has SWC has to do with thier success or failure for Rome? Rome was VC funded, afaik, turbopack still persists despite it being effectively obsolete with vite

Rome was going to be a whole toolchain built completely in Javascript. That was why they started with linting and formatting because they first had to implement the Javascript language parser and AST that was going to be the core of Rome. Linting and formatting was two easy things to start with while the ironed out the parser and AST implementation.

But when SWC/Esbuild came around to prove how much faster and efficient the compilation can be when written in a more performant language like Rust or Go, all hype around Rome died because no one wanted another Javascript toolchain.

Turbopack is backed by Vercel who obviously have a lot more capital behind them, but it's also built on Rust and at least has the potential to compete.

Reply on news.ycombinator.com