Next.js、React、TypeScript、前端工程化
This collection features articles and community discussions in the Web 开发 space, focusing on real-world engineering scenarios, tool selection, practical paths, and risk boundaries.
The page is continuously updated and serves as a learning roadmap, technology selection reference, and technical review entry point.
Feed systems are a demanding design problem: high read and write volume, ranking, real-time delivery, and scalability. This article explains push, pull, and hybrid models; timeline storage; popular content; pagination and end-of-feed behavior; and what to choose at tens of thousands versus hundreds of millions of users.
This is not about grinding coding problems; it is about writing better code. Learn which data structures matter most in production, including hash tables, trees, heaps, and graphs, how to analyze complexity, how to trade time for space, and why real skill means applying the right structure to the right engineering problem.
Choosing the wrong API style makes late-stage changes impossible. This article clarifies the philosophies and fit of REST (resources/universal/simple), GraphQL (flexible fetching/frontend-driven), and gRPC (high performance/internal microservices), plus scenario-based selection and REST's own pitfalls (versioning/status codes/nesting).
Writing Express routes is not enough for production Node.js. This article covers crash-safe error handling, logging and observability, configuration and secret management, database connection pools, scheduled jobs and queues, process management with PM2/cluster, and the realities of Node's single-threaded concurrency.
Ownership is Rust's hardest and most fundamental concept, enabling memory safety without a garbage collector. This article explains ownership, borrowing, and lifetimes in plain English, along with the common situations where beginners end up "fighting the compiler," to help you clear Rust's first and steepest hurdle.
Without the OWASP Top 10, Web applications are easy targets. This article explains XSS, SQL injection, CSRF, SSRF, insecure deserialization, and broken access control, including how they work, how modern frameworks defend against them, and the core habits of distrusting input, escaping output, and using parameterized queries.
State management is a perennial React topic, and too many choices can create confusion. Learn the right layer for useState, useReducer, Context, Zustand, and Redux; client vs. server state with React Query/SWR; and when Context or a global store is appropriate.
Still centering with float and margin? Modern CSS has moved on. This article covers the division of labor between Flexbox (1D) and Grid (2D), modern centering/equal-height/holy-grail patterns, Container Queries (component-level responsiveness), and new CSS capabilities (@layer/:has/logical properties)—so you write stronger layouts with less code.
Calling setState on every token off the network multiplies React commits. This article constrains only the frequency and freshness from event to commit, using per-frame coalescing, Event Timing, and LoAF to build release gates.
Mock tests all green, the real API may still have changed enums, defaults, or error semantics. This article brings specs, producer implementations, and test doubles into one versioned manifest, using bidirectional consistency and counterexamples to block "self-consistent but wrong" frontend tests.
Hooks are easy to use, but details around useRef, useMemo, and dependency arrays trap many developers. This article explains the Hooks mental model of closures and dependencies, common pitfalls such as stale closures and infinite loops, custom Hook extraction principles, and why useEffect is not a lifecycle method.
A page restored via the back button is not a normal reload; the JS heap and stale state can come back as-is. This article uses a pagehide/pageshow contract, resume generations, and a server operation receipt to prevent duplicate submissions and the continued use of stale permissions.