Our Predictions for 2023

Episode Summary

Executive Summary: Wes and Scott predict 2023 will be defined by server-side and HTML-first JavaScript patterns, stronger TypeScript inference, maturing runtimes like Deno and Bun, and more browser/runtime standards (View Transitions, container queries, subgrid, CSS :has). They also expect Rust-based tooling, more WASM and AI, and a growing shift away from React-centric thinking toward alternative frameworks like SvelteKit, Remix, Solid, and Qwik.

Main Topics: HTML-first and server-side JavaScript becomes the default: They expect SSR, progressive enhancement, and browser/web-API-driven patterns to keep gaining ground, with frameworks leaning more on HTML and forms rather than client-heavy SPA behavior. TypeScript inference and type safety improve developer ergonomics: The hosts argue inferred types, schema-driven typing, and generated types will become mainstream because they reduce duplication and make libraries feel magically type-safe. JavaScript runtimes mature: Deno, Bun, and Node catch up: They predict Deno will become more practical, Node will gain native TypeScript support, and Bun will become a legitimate deployment target for libraries and frameworks. Browser and TC39 proposals ship into daily workflows: They discuss likely progress on decorators, array non-mutating methods, temporal, record/tuple, pipeline operator, and other proposals that will shape code style and APIs. Tooling shifts toward Rust, faster linting, and faster type checking: They foresee Rust-based tools like STC, faster ESLint alternatives, and better performance in build/lint/typecheck workflows as codebases scale. Framework landscape broadens beyond React: The hosts expect more adoption of SvelteKit, Solid, Qwik, Remix, and other alternatives, driven by dissatisfaction with React’s complexity and by innovations like resumability. More platform features: View Transitions, container queries, subgrid, CSS :has: They highlight upcoming CSS and browser capabilities as major quality-of-life improvements that make apps feel more native and reduce custom JS work.

Key Arguments: Server-rendered, progressive-enhancement-friendly apps will keep growing because they better align with browser primitives and reduce client-side complexity. React’s architecture, especially its avoidance of the DOM, makes it harder to embrace web-native patterns like forms and browser APIs. TypeScript inference can eliminate repetitive type definitions and make complex libraries feel effortless to use. Deno is appealing because it provides a smoother out-of-the-box TypeScript experience than Node. Node will likely add built-in TypeScript support, reducing the current ecosystem friction around ts-node, tsx, loaders, and transpilers. Rust-based tooling will become more common because performance matters for large TypeScript codebases and developer feedback loops. Browser standards like container queries, subgrid, and View Transitions will materially change how developers build responsive and interactive interfaces. Alternative frameworks are gaining traction because they solve pain points React users are increasingly noticing. WASM and Rust will see more practical adoption where JavaScript is too slow for compute-heavy or media-heavy tasks. AI and machine learning will become more embedded in developer workflows and application features, beyond just code completion.

Data Points: SvelteKit adoption growth forecast: 5x increase - Scott predicts SvelteKit usage will grow from roughly 130,000 to about 700,000 on npm trends in 2023. SvelteKit current npm trend level mentioned: 130,000 - Used as the baseline for predicting future growth. SvelteKit starting npm trend level mentioned: 28,000 - Referenced as the earlier level before growth to 130,000. Firefox container queries support: 109 - They note Firefox shipped container queries in version 109. Safari subgrid support timing: September - They note Safari shipped CSS subgrid in September, just a couple of months before recording. GitHub tab width default: 8 spaces - Wes mentions GitHub’s default rendering for tabs as eight spaces. React Beta docs timeline: 5-year dev cycle - They joke that the React beta docs have been in development for roughly five years. Node TypeScript support timeline: by the end of the year - Wes predicts Node will run TypeScript without a plugin by year-end. Bun/different runtimes support expectation: by the end of the year - They expect more framework/runtime recommendations and deployment target decisions by year-end. Headphone purchase price: 60 CAD - Scott’s kid-friendly Wise noise-canceling headphones were about 60 Canadian dollars on sale.

Pivotal Quotes: "Server side, JavaScript sites become more the norm." — Scott Tolinsky: Opening prediction about the continued shift toward SSR and HTML-first app architecture. "Well, ain't that nice." — Wes Boss: Reaction to using 100% inferred TypeScript with little or no explicit type annotations. "The only way. If you're out there and you're saying, Scott's definitely going to do this, that's the only way I'm going to do this." — Scott Tolinsky: Joking about switching to tabs and humorously setting conditions for adopting the style preference.

Implications: Developers should expect more browser-native patterns, better TypeScript ergonomics, and faster toolchains. Teams that embrace web standards, edge-friendly code, and alternative frameworks may ship simpler, faster apps with less JavaScript overhead.