May 7, 2024

Rollup: improved tree-shaking due to tracking of const parameter values

github.com github.com/liuly0322 @lukastaegert@webtoo.ls @rollupjs@webtoo.ls

Rollup now tracks which values parameters receive: If a parameter always receives the same value, dead code inside functions can sometimes be eliminated. The first comment lists three REPL examples that demonstrate what is and isn’t possible.

Node v22.1.0: automatic on-disk code caching

nodejs.org @targos@fosstodon.org @nodejs@social.lfx.dev

“When [enabled], whenever Node.js compiles a CommonJS or ECMAScript Module, it will use on-disk V8 code cache persisted in [a] directory to speed up the compilation. This may slow down the first load of a module graph, but subsequent loads of the same module graph may get a significant speedup if the contents of the modules do not change. Locally, this speeds up loading of test/fixtures/snapshot/typescript.js from ~130ms to ~80ms.”

Live Workshop: Sentry 🤝 Supabase: Find the slow queries in your database

sentry.io Sponsor

Learn how to improve the performance of your PostgreSQL database and get real-time error alerts at our upcoming workshop May 16th. Plus join us live and you’ll have a chance to win swag.

Deno 1.43: improved Language Server performance

deno.com github.com/lambtron github.com/bartlomieju @marvinh@infosec.exchange github.com/iuioiua @deno_land@fosstodon.org

“[...] we have enhanced the performance of Deno in IDEs by reducing auto-completion times from 6-8 seconds to under one second in large codebases and significantly decreasing memory usage.”

Vitest 1.6.0: faster “forks” pool, comparing benchmark results, and more

github.com @vitest@webtoo.ls

Highlights (quoting the mastodon post):
  • Performance improvements for the “forks” pool
  • --standalone flag to start Vitest in watch mode without running tests
  • --compare flag to compare benchmark results
  • New options to inject scripts in browser mode

npm feedback migrates from discussions in a GitHub repository to GitHub Community

github.blog 

nrr: minimal, fast npm script runner

github.com github.com/ryanccn

Code health tip: refactor code with complex boolean expressions to make it easier to understand

testing.googleblog.com 

URL.parse(): new URL() without exceptions

kilianvalkhof.com github.com/kilian

JavaScript language design and implementation in tandem

dl.acm.org github.com/sukyoung github.com/jhnaldo

In this article, we present how to automatically extract a mechanized specification from a prose specification and how useful it can be in practice.

Using the example of JavaScript, we show how mechanized specifications can be used to detect conformance bugs between language specifications and existing JavaScript engines in major Web browsers, and to generate more special-purpose JavaScript implementations, such as static analyzers, in a correct-by-construction manner.

We propose a new approach to programming language development as a promising direction for the future: first design the language in a mechanized specification and then generate both human-friendly specifications written in diverse natural languages and correct-by-construction implementations and tools from the mechanized specification.

‍

Poll: What is your preferred JavaScript bundler (not necessarily the one you use the most)?

We’ll share the results in one of the next editions!
‍

This email was sent to {{ email | default }}. You can unsubscribe from this list here or update your preferences.

‍