Official Ctrip Technology account, sharing and discussing growth.
This article presents a Taro plugin that leverages cssModules syntax to shrink compiled style files by shortening class names and de‑duplicating property values, achieving a 50‑70% reduction in CSS size while preserving the original development experience for React‑based mini‑programs.
This article presents a comprehensive case study of Ctrip's train ticket SMS recall business, detailing the design, implementation, and evaluation of response‑based conversion rate models and uplift models to improve marketing ROI through causal inference and machine‑learning techniques.
This article explains how to integrate the Baidu map plugin into a FlutterBoost hybrid project, describes the PlatformView mechanism, analyzes rendering and thread‑synchronization details, and provides concrete solutions for iOS white‑screen, Android freeze, memory‑leak, custom markers and visible‑range problems.
This article presents a comprehensive design and implementation of a serverless delayed message queue on AWS, evaluating various solutions such as RabbitMQ, ActiveMQ, RocketMQ, Redis, and ultimately proposing a hybrid Kafka‑SQS‑DynamoDB architecture with performance optimizations and operational insights.
The article details how Ctrip's train ticket business revamped its multi‑engine OLAP metrics platform by consolidating to the StarRocks MPP database, describing the new architecture, query workflow, data synchronization strategies, practical lessons, and the resulting dramatic improvement in query latency and reliability.
This article introduces the fundamental concepts and theoretical frameworks of causal inference, explains Rubin's potential outcomes and Pearl's causal graph models, and demonstrates their practical deployment through uplift modeling, propensity‑score matching, synthetic control, and regression‑discontinuity case studies within Ctrip's train ticket business.
This article explains how to improve CSS animation performance by understanding the browser rendering pipeline, using GPU‑accelerated transforms, avoiding costly properties and complex selectors, leveraging will‑change and requestAnimationFrame, and preferring CSS over JavaScript for smoother, lower‑latency visual effects.
This article examines common timeout problems in microservice architectures, identifies root causes such as connection and socket timeouts, and presents ten practical optimization techniques—including setting appropriate timeouts, rate limiting, cache improvements, thread‑pool tuning, GC and JIT adjustments, NIO async programming, host migration, and network checks—to enhance system stability and performance.
This article describes a cross‑platform solution that enables a single set of Web component code to be shared across Native, React Native, and mini‑program environments by leveraging Web Components, environment‑aware builds, and bridge communication, thereby reducing development effort and accelerating release cycles.
This article presents a graph‑neural‑network driven, unsupervised approach that builds heterogeneous user‑feature graphs, learns node weights, constructs user‑user similarity graphs, and applies threshold‑based clustering to identify abnormal registration clusters for fraud detection in Ctrip's business travel platform.
