TypeScript
AdoptLanguages & Frameworks
A typed superset of JavaScript used for building application code.
Why it's here
Placed in Adopt: 14 article(s) of evidence from 4 source(s), led by research-stage coverage, with 6 in the last 30 days. Confidence 80%.
Evidence (14)
- 7InfoQ·8/10/2026framework_updateAngular v22 adds stable Signal Forms and default OnPush change detection
Google has released Angular v22 with several framework updates, including the stabilization of Signal Forms, OnPush change detection by default, and a new experimental WebMCP integration for AI-related tooling. The release also introduces a new @Service() decorator, supports TypeScript 6, and removes deprecated features.
- 8InfoQ·8/3/2026framework_updateMicrosoft Ships TypeScript 7.0 with Native Go Compiler
Microsoft has released TypeScript 7.0 with a native compiler written in Go, delivering build speeds that are reported to be 8x to 12x faster in real-world codebases. The release does not yet include a stable programmatic API, which is expected in TypeScript 7.1, and a compatibility package is available for existing tooling.
- 4Martin Fowler·7/30/2026researchRefactoring Cuts Agentic Coding Costs
A Thoughtworks CTO describes an experiment refactoring a 17,155-line Rust data access layer built by coding agents. By measuring token and time costs for the same change before and after each refactoring step, the article argues that cleanup work can reduce the cost of future agent-generated changes.
- 6Hacker News·7/26/2026open_sourceVercel releases Scriptc, a TypeScript-to-native compiler
Scriptc is a Vercel Labs project that compiles TypeScript to native binaries without bundling a JavaScript engine. The repository presents it as an experimental tool aimed at producing smaller, faster standalone executables for server-side use. Community discussion on Hacker News focused on its design, performance tradeoffs, and developer experience.
- 8Hacker News·7/15/2026researchStar Fleet Claims 20 Erdős Problem Solutions in Parallel
Star Fleet, a Lean 4-based AI math system, claims it solved 20 Erdős problems using up to 20 parallel agent harnesses, each running a separate GPT-5.6 instance. The setup combines large-scale compute, theorem search, proof verification, and long-term dependency tracking to tackle open mathematics problems.
- 6InfoQ·7/14/2026framework_updateGoogle Adds Agents API to Genkit for TypeScript and Go
Google has released the Genkit Agents API in preview for TypeScript and Go. The open-source update adds a unified chat() interface with message history, tool loops, streaming, and state persistence, plus detached turns and interruptible tools for human-in-the-loop workflows.
- 8Simon Willison·7/8/2026framework_updateBun rewritten in Rust with agentic coding workflows
Bun’s maintainer describes a large-scale rewrite from Zig to Rust, using coding agents, automated trial runs, and adversarial review to port a million-plus lines of code. The rewrite was enabled by a TypeScript test suite used as a conformance harness, and the new Rust version is already running in Claude Code with modest performance gains on Linux.
- 9Hacker News·7/8/2026framework_updateTypeScript 7 launches with a native Go port and major speed gains
TypeScript 7 is now available as a native port built in Go, with the project claiming 8x to 12x faster full builds and lower memory use than TypeScript 6. The release also improves editor features through LSP support and faster multithreaded performance across common tools such as VS Code and Visual Studio.
- 3InfoQ·7/6/2026researchPractical Robustness in Rust Beyond Memory Safety
Andy Brinkmeyer presents ways engineering leaders and architects can use Rust to build more robust systems beyond basic memory safety. He highlights how ownership, enums, and the typestate pattern can move complex runtime protocols into compile-time checks and reduce entire classes of bugs.
- 4Hacker News·6/30/2026open_sourceParsing Instead of Validating in TypeScript
The article argues that TypeScript code often validates data repeatedly instead of parsing it into more specific types that preserve guarantees. It discusses using branded or nominal-style types to encode parsed values like email addresses so later code can rely on the type system instead of re-checking conditions.
- 5Martin Fowler·5/27/2026researchTest suite as a regression sensor for AI code
The article describes using a set of automated “sensors” to monitor maintainability in AI-generated codebases, with the test suite serving as a regression sensor. It discusses combining tests with static analysis, dependency checks, and mutation testing to catch issues early and support self-correction before changes reach humans.
- 4Martin Fowler·5/20/2026researchThree More Static Code Analysis Sensors
Martin Fowler highlights a practical follow-up on using static code analysis and related sensors to help coding agents keep codebases maintainable. The article discusses feedback tools such as linting, dependency rules, coupling analysis, modularity review, and mutation testing to catch issues early and support self-correction.
- 5Martin Fowler·5/19/2026researchMaintainability sensors for coding agents
The article describes practical experiments with using automated “sensors” to help AI coding agents keep a codebase maintainable. It highlights tools such as linting, dependency checks, test coverage, mutation testing, and structural analysis to catch maintainability issues early and support self-correction before changes reach humans.
- 5Martin Fowler·4/2/2026researchHarness engineering for coding agent users
The article proposes a harness engineering mental model for coding agents to improve trust in AI-generated code. It argues for combining feedforward guides and feedback sensors, using both deterministic tooling and semantic evaluation to reduce errors before human review.