Blog

N-Queens separates candidates who know backtracking from those who understand constraint propagation. The O(1) diagonal check using sets is what interviewers want to see.

The duplicate-skip pattern is what separates Combination Sum I from II. Master sort + skip, in-place grid marking for Word Search, and why `i > start` matters.

Most teams hit 'use client' everywhere and wonder why their bundle explodes. The fix is a single mental model: keep the boundary as low as possible, and use composition to thread server-rendered content through client islands.

Build a spec-compliant Promise from scratch in 75 lines of JavaScript. Understand state machines, microtask scheduling, thenable resolution, and why .then(s,e) and .then(s).catch(e) aren't the same.

BST problems aren't just binary tree problems — the sorted-order property changes everything. Master validate, kth smallest, and the iterator pattern that shows up in real interviews.

React Context is great for static config but a footgun for dynamic state. Learn how to split contexts by update frequency and separate state from dispatch.

Kafka and RabbitMQ solve different problems. Pick the wrong one and you'll either miss replay when you need it, or run a cluster for work that a queue handles fine.

Three linked list problems, one underlying playbook: dummy heads, divide-and-conquer merging, and the find-middle/reverse/interleave pipeline that makes LC 143 tractable.

One BFS template, three interview problems. Level-order, zigzag traversal, and right side view all reduce to the same queue loop with a small variation each time.