TLDR Web Dev

TLDR WEB DEV 2024-10-16

6 sections17 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01GENERAL

02ARTICLES & TUTORIALS

WHY YOU SHOULDN'T FORGET TO OPTIMIZE THE DATA LAYOUT

The way data is organized in memory (data layout) can impact a program's performance. Choosing the right data layout, such as Array of Structures (AoS) or Structure of Arrays (SoA), can optimize cache line access and leverage vectorized instructions (SIMD) for faster execution. SoA, for example, can be much faster for reading operations, especially when combined with lightweight compression techniques like frame-of-reference encoding.

cedardb.com12 min read

LOCALSTORAGE VS. INDEXEDDB VS. COOKIES VS. OPFS VS. WASM-SQLITE

This article compares different methods for storing data in web browsers: Cookies, LocalStorage, IndexedDB, OPFS, and WASM-SQLite. Each method is analyzed for their features, limitations, and performance in various scenarios, including write and read latency for single and bulk operations.

rxdb.info20 min read

PYTHON: RANGE IS NOT AN ITERATOR

Iterators and range objects are different in Python. While both are considered "lazy iterables" as they compute values on demand, they differ in their underlying mechanisms. Iterators are stateful, meaning once an item is consumed, it's gone. Range objects are not iterators - they are "lazy sequences" that can be looped over repeatedly without exhausting their values.

treyhunner.com8 min read

03OPINIONS & ADVICE

SMURF: BEYOND THE TEST PYRAMID

SMURF (Speed, Maintainability, Utilization, Reliability, and Fidelity) is a framework for better testing. Prioritizing one factor can have implications for others, so it's best to have a holistic approach when writing tests for applications.

testing.googleblog.com3 min read

SCALE RUINS EVERYTHING

The pursuit of unicorn status by venture capital funds has led to the creation of companies that, while initially seemingly harmless, become detrimental at scale. This is because the need for massive growth drives companies to exploit resources and disrupt markets. The recent decline in venture capital funding offers an opportunity for more non-hyper growth companies that are still sustainable businesses.

coldwaters.substack.com6 min read

A MISSING IDE FEATURE

This developer proposes a missing IDE feature: automatically folding method bodies by default. This simple change improves code readability, as function signatures are often more important than their bodies.

matklad.github.io6 min read

04LAUNCHES & TOOLS

REACT NATIVE BOTTOMSHEET V5 IS FINALLY OUT!

React Native Bottom Sheet v5 has been released. It features a complete gesture solution rewrite, support for React Native Web, dynamic sizing, FlashList integration, and performance improvements.

gorhom.dev3 min read

KAIBANJS

KaibanJS is a JavaScript framework for building and managing multi-agent systems.

github.comGitHub repository

FTWA

FTWA allows users to turn any website into a lightweight, fast, and visually appealing app. It removes the need for browser frames and creates a native app-like experience.

ftwa.mathix.devTool

05MISCELLANEOUS

WHY TECHIES LEAVE BIG TECH

Software engineers and tech leaders are leaving prestigious Big Tech companies. There are various reasons why, like the decline in job security due to recent layoffs, the allure of professional growth and leadership opportunities offered by startups, and the perceived limitations of career progression within large corporations. There's also the impact of equity compensation and the potential for significant pay drops when stock prices fall or initial grants expire.

newsletter.pragmaticengineer.com31 min read

META'S OPEN AI HARDWARE VISION

Meta has presented its open AI hardware vision, showcasing new designs for AI platforms, racks, network fabrics, and components. To support its large-scale AI workloads, Meta has built high-performance training clusters with thousands of GPUs. To address the increasing power and bandwidth demands of AI, Meta has Catalina, a new high-powered rack designed for AI workloads, and Grand Teton, a next-generation AI platform now supporting AMD accelerators.

engineering.fb.com7 min read