Hacker News Reader

Comment by Helianthus | Hacker News Reader

>As for traction, if Dart doesn't get any traction (after it's actually released, and had a good chance to succeed of course!), inside Google or outside, then it probably _should_ be abandoned, right?

Yep. So goes the argument. When/how is it going to be released in a meaningful way, such that it can 'actually' gain traction?

People are using CoffeeScript and Typescript in production right now. Dart has been a theoretical project for far too long.

>The Dart and Chrome teams firmly believe that the adoption of Dart will entirely depend on it's success as a compile-to-JS language.

Yep. It's not doing so well.

>The FUD around Dart is just incredible.

Maybe it's not FUD, maybe it's legitimate criticism.

In any case, I wish you good luck.

View on HN

Dart 1.0 will be released pretty soon, as Lars & co, have said. It takes a long time to create a new language and core libraries from scratch.

But I can't believe you complain that it's not released yet, and that it doesn't have much traction in the same post...

I think Dart adoption is going along just fine, considering it's pre-release state for the last two years. If too many people had already starting using it, they might have had unmeetable expectations about it's stability and readiness. Better that there are fewer early adopters who are aware of and willing to take the pain of the evolving platform.

>But I can't believe you complain that it's not released yet, and that it doesn't have much traction in the same post...

Where's the contradiction?

For one, other projects have gotten traction LONG before the stable release. Mongo and Node.js come to mind.

Second, the complaint can be stated as: "If you want it to gain traction, why do you keep tinkering on it for such a long time without a stable release"?

Damned if you do and damned if you don't. If they release it and iterate on it for years in the public before official release, you complain about adoption or lack of users.

If they were to bake it internally first until it is stable and then release it "done", then everyone complains about a closed process.

And do you have any numbers to compare Dart vs Typescript vs CoffeeScript adoption?

>If they were to bake it internally first until it is stable and then release it "done", then everyone complains about a closed process

Well, Go did just that, and I didn't see anyone complaining.

If they're baseless hypothetical assertions laced with strawman arguments from irrelevant technologies, I'd tip towards FUD.

I ain't sure if that's directed at me, but assuming it is, CoffeeScript and Typescript and, heck, even asm.js are direct competitors to Dart. It's all transpiled languages competing for mindshare. Dart isn't doing well on that field.

There's nothing regarding fear, uncertainty, or doubt in that assertion.

Was directed at the general FUD towards Dart, not your comment.

That said nothing like Dart exists, which is a ground-up new implementation of a new structured web language to build and maintain high-performance complex web apps. There are many parts to Dart, like the Dart SDK which provides a set of well-written and consistent APIs, smoothing over the DOM and browser quirks that also introduces a new Future/Streams composable async model simplifying IO and handling events in js web apps.

The dart2js component is what Dart shares with other transpiled languages like CoffeeScript, TypeScript and asm.js. Out of this only TypeScript shares the goal of a structured/toolable language for building complex web apps, which is essentially just an optionally-typed superset of JS that's forward-looking to features proposed by the ECMA standards committee. Since TypeScript is a tooling/transpiler that sits on top and is dependent on JS it will have a slower dev/edit/reload cycle since it always needs to compile before running, its tied to JS semantics so it wont be able to have the same level of advanced tree-shaking Dart can and as its performance is tied to JS VMs it wont be able to take advantage of DartVM "Snapshots" for instant start-up times or the 2x faster predictable runtime performance of the DartVM.

Dart is also the only strategy that allows us to get a "clean-break" from JS. Personally I think it'd be a tragedy if in 10 years time everything still depends on and we can't innovate past JS, killing any hope being able to natively run other languages in browsers. I strongly believe in order to progress we need to re-invent, to tear things down, keep what's good and discard the warts and legacy, not build on top of it. Maybe it turns out Dart isn't it, but I sure hope something better succeeds past JS.

> that also introduces a new Future/Streams composable async model simplifying IO and handling events in js web apps.

Sounds like RxJS to me.

Which is tacked onto JS callbacks and not engrained into every IO operation like it is with Dart.

No language really competes pre-release

> Maybe it's not FUD, maybe it's legitimate criticism.

"If I could take a clean sheet of paper and write [a new language] that retains all the goodness of [JavaScript] ... I would not have come up with anything like Dart." -- Douglas Crockford

How is this legitimate criticism? Especially since Dart has all the goodness of JavaScript except monkey patching (or is that one of the good parts?)

What frustrated me about Douglas Crockford's response is that he didn't come out and say what he'd like a JavaScript replacement to look like. That would have been interesting.

That was my thought too. He said recently in a video interview the he wished JavaScript had tail call optimization, so it would probably be something along functional programming lines.

I think it was this one, but they don't publish a transcript so I can't easily verify:

http://channel9.msdn.com/Events/Build/2013/9-011

He's been very clear about what he wants as a replacement for javascript for many years, hiding it in plain sight in a little book called "Javascript the good parts" and codifying it in JSLINT. It just looks too much like Javascript to be recognised as a replacement.

To be fair, people use Dart in production right now, too. In fact, there are high profile Google projects being written in Dart (unfortunately, they aren't released yet). See Blossom's post for one example: http://www.ramen.io/post/46936028144/we-are-switching-to-dar...

Reply on news.ycombinator.com