whoway Technical Notes

A Monotonous Queue maintains elements in a monotonically increasing or decreasing order. It allows deletions from both ends but insertions only at the tail. A key property is that each element enters and leaves the queue exactly once.

Common applications include:

• Sliding Window Maximum/Minimum (e.g., LeetCode 239).
• Finding the minimum/maximum in a sliding window of a sequence.

Sorting algorithms are categorized into five main types based on their underlying logic. Understanding these is critical for solving algorithmic problems (e.g., LeetCode 912) without using built-in functions.

• Insertion-based: Direct Insertion Sort, Binary Insertion Sort, Shell Sort.
• Selection-based: Simple Selection Sort, Heap Sort (focuses on the Heap data structure).
• Exchange-based: Bubble Sort, Quick Sort (focuses on dual-pointer and divide-and-conquer techniques).
• Merge-based: 2-way Merge Sort (focuses on space-time trade-offs).
• Non-comparison based: Radix Sort, Counting Sort, Bucket Sort.

The evolution of web development is characterized by how requests enter a program, how programs persist in memory, and how resources are reused. This evolution is divided into three main stages:

1. CGI (Common Gateway Interface): The starting point where the Web Server forks a new process for every request to execute a script.
2. Containerization (FastCGI / Servlets): Introduction of long-running processes or containers using thread pools to handle requests, reducing the overhead of process creation.
3. Application as Server: The application, framework, and runtime are bundled together (e.g., Spring Boot, Gin), where the application itself acts as the server.

The VuePress theme-hope layout consists of global elements and page-specific components that can be customized via theme options and page frontmatter.

Global Layout Elements:

• Navbar
• Sidebar
• Footer

Page-Specific Elements:

• Breadcrumbs
• Page Title and information
• Table of Contents (TOC)
• Meta information (update time and contributors)
• Comments section

Interface Elements:

• Dark Mode toggle button
• Back to top button
• Print button

Most AI application technologies based on Large Language Models (LLMs) can be distilled into two fundamental components: Prompt Engineering and Context Engineering.

• Prompt Engineering: The art of how to 'ask' the AI. It determines the starting point and direction of the AI's reasoning.
• Context Engineering: The process of 'feeding' and 'managing' the AI's memory. This is critical for achieving professionalism and continuity in specific tasks, involving the organization, retrieval, updating, and limitation of background information within the AI's context window.

The evolution of dynamic web services can be categorized into three distinct architectural paradigms based on their core subject, deployment unit, concurrency model, and isolation method:

• CGI: The Web Server is the core subject. It deploys scripts and uses OS process isolation with a process-based concurrency model.
• Servlet: The Container is the core subject. It deploys WAR files and uses ClassLoader isolation with a thread-pool concurrency model.
• Application as Server: The Application itself is the core subject. It deploys as a JAR or binary and uses process/thread/coroutine-level isolation with a thread or coroutine concurrency model.