AI summary by TwinMind

JavaScript might become two languages (and it's dramatic)

Theo - t3․gg

24m 33s

JS Sugar/JS Zero Proposal Overview

  • New TC39 proposal to split JavaScript into two parts
    • JS Zero: core language implemented by browsers/engines
    • JS Sugar: syntax features implemented only by build tools (Babel, webpack, etc.)
  • Presented by Google, Mozilla, Apple, and Sony at TC39 meeting
  • Proposal argues build tools already widely adopted, should leverage this for faster language evolution
  • Would defer syntax features to tooling authors instead of browser implementers

Core Arguments from Implementers

  • JavaScript engines face increasing complexity burden from new features
    • Security vulnerabilities in V8 from: BigInt, class scopes, built-in promise subclassing, resizable buffers, weak refs, logical assignments, top-level await
    • Performance issues: ESMs bad for networking, iterator proposals hard to optimize, too many allocations
    • Stability concerns: enterprises freeze browser versions, nullish coalescing broke Firefox
  • Users (not developers) should be primary customer of browser implementations
  • New language features provide minimal user benefit while increasing security/performance risks
  • Complexity should live "at the edge" (in tools) rather than in widely-used runtimes

Proposed Implementation Structure

  • Current flow: TypeScript/modern JS → build tools → standard JS → browser execution
  • Proposed flow: TypeScript + JS Sugar + JS Zero → build tools → JS Zero → browser execution
  • JS Sugar features would be standardized but never implemented in browsers
  • Tools would compile JS Sugar down to JS Zero for browser compatibility
  • Future features could start as JS Sugar and potentially move to JS Zero if warranted

Community Backlash and Criticism

Key Critics

  • Yehuda Katz (Ember community legend): "RIP vanilla JS" - build tools would become mandatory
  • Brendan Eich (JavaScript creator): implementers lack real web development experience, shouldn't call the shots
  • Canada Honk: tooling ecosystem already unstable, lacks resources for this responsibility

Major Concerns Raised

  • Tooling Resource Constraints
    • Current tools (Babel → SWC → OXC) already in constant flux
    • Rewriting tools in Rust while adding new feature burden would slow progress significantly
    • Legacy tech debt in tools from previously implemented features
  • Developer Experience Impact
    • Developers would need to understand two different JavaScript languages
    • Browser console wouldn't support JS Sugar syntax
    • Similar to requiring understanding of V8 bytecode - unnecessary complexity
  • Ecosystem Fragmentation
    • Server-side runtimes (Node, Deno, Bun) would need to implement JS Sugar features
    • Risk of features being implemented differently across tools
    • Potential for tools to hide unfixed mistakes longer

Technical Implementation Concerns

  • Feature Stability Issues
    • JS Sugar could undermine TC39's staging process designed for stability
    • Risk of bad features advancing further before being caught
    • Tools might not upgrade consistently, leaving developers behind
  • Node.js/Server-side Impact
    • Async generators commonly used in Node but claimed underutilized by browser implementers
    • Server-side runtimes would bear implementation burden instead of underlying engines
  • Security Implications
    • If browser engines struggle with security in new features, how will smaller tool teams handle it?
    • Potential for more vulnerabilities in distributed tooling ecosystem

Alternative Solutions Mentioned

  • ESM bundling proposal for concatenating files (addressing networking concerns)
  • Better communication between implementers and developer community
  • Improved staffing for JS VM teams to include developer advocates
  • Source map improvements (should happen regardless)

Industry Position Split

  • Supporters: Malte Ubl (Vercel CTO, former Google) sees this as opposite of web components strategy - good direction
  • Opposition: Broad resistance from prominent community figures
  • Neutral/Concerned: Many acknowledge implementer burden but question if this solution helps developers or users

Fundamental Philosophical Divide

  • Implementers argue helping developers vs. serving users are conflicting goals
  • Critics argue this creates false dichotomy - helping developers ultimately helps users
  • Question of whether browser teams should focus purely on user-facing benefits vs. developer productivity
  • Debate over whether complexity should be handled by specialized tools vs. standardized in browsers

TwinMind

Get summaries like this for your own meetings

TwinMind takes notes in the background during meetings, lectures, and calls, then writes the summary for you.

Free plan to get started

No bot joins your calls

Phone, desktop & Chrome apps

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome