Skip to content
Permalink
Branch: master
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
95 lines (55 sloc) 7.04 KB
title
Frequently Asked Questions

Frequently Asked Questions

What is One JavaScript?

One JavaScript, or 1JS for short, means avoiding splintering JavaScript by making each version backwards-compatible with the previous versions. We evolve the language by adding new API's and new syntax. We don't change the behavior of existing features, and we don't add new modes or versions.

Why don't we remove or fix bad features?

If we did, existing web content would break. No browser vendor is willing to be the first to make such changes, because a rational end user perceives them as browser bugs and switches to a different browser.

The only time we can make an incompatible change is if the web happens not to depend on it. There's no way to know for sure whether this is the case except by experimenting. Browser vendors can take a chance by making an incompatible change in alpha or beta versions of their browsers and seeing whether they get any bug reports about breaking content.

Why don't we make breaking changes with a language version?

Opt-in versioning has a bad track record on the web. VBScript, JavaScript 1.2, and XHTML are good examples.

Versions or modes may sound appealing, but they have the wrong defaults: an unspecified version (e.g., in a script tag or an inline attribute) always gets the old version—to get an improved language you always have to specify a version. It's not even possible to specify a language version in an inline HTML attribute. Maybe worst of all, adding modes or versions adds a dimension to the platform matrix for application testing and tracking platform feature availability.

For all these reasons, opt-in versioning faces an uphill adoption battle. It's also a huge implementation and testing burden on platform vendors—it increases their set of sub-platforms exponentially—so browser vendors typically oppose new modes.

Why don't we deprecate features?

Deprecation doesn't work on the web. Since we can't remove bad features, a developer has little incentive to stop using a feature just because somebody somewhere doesn't like it. Labeling features "deprecated" without ever removing them is pointless.

How can we use new syntax on old browsers?

New syntax takes a while to reach enough users that developers can start using it. However, adding syntax to JavaScript has historically been successful, even if it's slow. An example is try/catch, which was introduced in ES3 (1999). It took years before widely-deployed applications could use it, but it's now a full-fledged part of the web platform.

A good approach to using new syntax on old browsers is to use a "transpiler" to compile new features to old versions of JavaScript. Google's Traceur compiler is an experimental project to do this for ES6.

When will the next version of JavaScript be available?

ECMAScript Edition 6 is targeting a 2013 spec release, but browsers are already implementing individual features.

Sites like Can I Use and HTML5 Please track the release and availability of features on the web platform. Hopefully these sites or others like them can track individual features of ES6 as well.

Will we add new restrictions to strict mode?

No. Strict mode was a one-time event to fix some of the problems in ES3. Going forward, modules will automatically be strict, making it less necessary to opt in to strict mode explicitly to get these fixes. But the goal of One JavaScript is to avoid any more language fixes with modes.

Why don't we standardize a bytecode VM?

Bytecode languages often require a verifier which can hurt startup time, and lead to code bloat which can affect download footprint and network latency. JavaScript successfully serves as a target for compilers from other source languages, and one of the explicit goals of TC39 is to add features that help code generators in addition to programmers.

Classes? Is JavaScript becoming Java?

No way! The proposed classes for ES6 are a syntax for JavaScript's ubiquitous constructor-and-prototype pattern. Classes are completely dynamic—in fact, they're even an expression form, so you can very easily and expressively generate new classes at runtime.

How do we avoid design-by-committee?

TC39 uses the Harmony process for design work. Small teams work independently on a design sketch and present it to the rest of the committee for discussion. These teams act as the champions for their design. If the committee decides the design has become promising enough to be worth pursuing, the sketch is promoted to an official proposal. The champions of the proposal continue to lead the design work and refinements to the proposal.

Do all those design sketches mean JS will get bloated?

Definitely not. Most design sketches are not accepted as official proposals. The design process requires going through many, many ideas before picking the best ones. TC39 makes design sketches public to encourage as much community feedback as possible, as early as possible.

How can I suggest a feature?

Send an email to es-discuss!

I'm nervous about posting to es-discuss. Any advice?

Just keep it short and constructive. Everyone is welcome to contribute or provide feedback.

If you aren't sure exactly what the solution or design you have in mind should look like, don't worry: just describe the problem you perceive. Feedback from developers about common problems can be particularly valuable.

Also: please trim cited text to only the relevant parts (Gmail users: the UI hides cited text from you but it's still there—please unhide and trim!).

How do you spell JavaScript? ECMAScript? Ecma?

JavaScript, ECMAScript, and Ecma.

Don't ask.

You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.