2025-11-21 Tech Briefing | AWS optimizes CI/CD build efficiency

Friday

Welcome to this edition, featuring 30 articles.

📚 Contents/📝 TL;DR

📚 Contents

📝 TL;DR

Primary

Amazon Web Services Introduces Remote Build Cache in ECR to Accelerate Docker Image Builds

AWS optimizes CI/CD build efficiency

Amazon Web Services Introduces Remote Build Cache in ECR to Accelerate Docker Image Builds

AWS optimizes CI/CD build efficiency

From 'User Thinking' to 'Cultivation Thinking': Discussing the Arrogance and Bias of Tech Giants

Tech giants' shift from 'user thinking' to 'cultivation thinking' has sparked public outrage.

From 'User Thinking' to 'Cultivation Thinking': Discussing the Arrogance and Bias of Tech Giants

Tech giants' shift from 'user thinking' to 'cultivation thinking' has sparked public outrage.

AWS Introduces Landing Zone Accelerator General Configuration and LZA Compliance Workbook

AWS releases LZA general configuration and compliance workbook to simplify secure and compliant deployment.

AWS Introduces Landing Zone Accelerator General Configuration and LZA Compliance Workbook

AWS releases LZA general configuration and compliance workbook to simplify secure and compliant deployment.

Using Application Load Balancer Target Optimizer to Improve Application Performance

AWS ALB introduces Target Optimizer, precisely controlling backend concurrency to improve application performance.

Using Application Load Balancer Target Optimizer to Improve Application Performance

AWS ALB introduces Target Optimizer, precisely controlling backend concurrency to improve application performance.

AWS re:Invent 2025 Amazon EKS and Kubernetes Conference Guide

AWS re:Invent 2025 Kubernetes Conference Schedule Released

AWS re:Invent 2025 Amazon EKS and Kubernetes Conference Guide

AWS re:Invent 2025 Kubernetes Conference Schedule Released

Meituan LongCat Interaction Team Releases Technical Report on Large Model Interaction System WOWService

Meituan Releases Technical Report on Large Model Interaction System WOWService.

Meituan LongCat Interaction Team Releases Technical Report on Large Model Interaction System WOWService

Meituan Releases Technical Report on Large Model Interaction System WOWService.

No Data Upload Required! Apple Officially Open Sources Embedding Atlas, an Interactive Vector Visualization Tool for Desktop Using Rust+WebGPU

Apple opens sources Embedding Atlas, an interactive embedding vector visualization tool for desktop.

No Data Upload Required! Apple Officially Open Sources Embedding Atlas, an Interactive Vector Visualization Tool for Desktop Using Rust+WebGPU

Apple opens sources Embedding Atlas, an interactive embedding vector visualization tool for desktop.

Kiro is Now Available, Core Features Receive Enhancements

Amazon Kiro is now live, supporting property-based testing and terminal CLI interaction.

Kiro is Now Available, Core Features Receive Enhancements

Amazon Kiro is now live, supporting property-based testing and terminal CLI interaction.

Achieving High Availability with Block-Level Replication in Amazon RDS for SQL Server Web Edition

High availability achieved with SQL Server Web Edition

Achieving High Availability with Block-Level Replication in Amazon RDS for SQL Server Web Edition

High availability achieved with SQL Server Web Edition

Kiro 'Startup Support Program': AI Agent Helps You Scale from Idea to Production Without Worry

Kiro launches a startup support program, offering free AI Agent quotas

Kiro 'Startup Support Program': AI Agent Helps You Scale from Idea to Production Without Worry

Kiro launches a startup support program, offering free AI Agent quotas

Finding Direction in Platform Maturity

Don't just focus on tools; the core of platform maturity is eliminating cognitive load.

Finding Direction in Platform Maturity

Don't just focus on tools; the core of platform maturity is eliminating cognitive load.

Essential

Top Strategies for Building Highly Available Systems

Building highly available systems is the foundation of modern applications.

Top Strategies for Building Highly Available Systems

Building highly available systems is the foundation of modern applications.

Linus Talks About AI and Atmosphere Programming: It's Good for Beginners, But Not for Production

Linus states that AI is suitable for beginners but not for production code

Linus Talks About AI and Atmosphere Programming: It's Good for Beginners, But Not for Production

Linus states that AI is suitable for beginners but not for production code

Introducing Strands Agent SOPs – AI Agents with Natural Language Workflows

Standardizing AI agent workflows

Introducing Strands Agent SOPs – AI Agents with Natural Language Workflows

Standardizing AI agent workflows

Exclusive Share from a Top-Tier CIO: How AI is Reshaping Developers' Future for the Next Decade

AI will disrupt the developer role and enterprise efficiency metrics

Exclusive Share from a Top-Tier CIO: How AI is Reshaping Developers' Future for the Next Decade

AI will disrupt the developer role and enterprise efficiency metrics

E215｜Crypto Miners with Electricity Advantage: Will They Be the 'Gold Miners' of the AI Era?

Crypto miners leverage electricity advantage to transition into AI infrastructure, facing financial risks.

E215｜Crypto Miners with Electricity Advantage: Will They Be the 'Gold Miners' of the AI Era?

Crypto miners leverage electricity advantage to transition into AI infrastructure, facing financial risks.

Amazon S3 Introduces Attribute-Based Access Control for General Purpose Storage Buckets

AWS S3 introduces attribute-based access control, enabling automatic permission management via tags.

Amazon S3 Introduces Attribute-Based Access Control for General Purpose Storage Buckets

AWS S3 introduces attribute-based access control, enabling automatic permission management via tags.

Dark Data Tax: How Stockpiling Poisons Your AI

Data stockpiling causes data obesity and AI poisoning, focusing on data quality rather than quantity.

Dark Data Tax: How Stockpiling Poisons Your AI

Data stockpiling causes data obesity and AI poisoning, focusing on data quality rather than quantity.

What New Innovations Have Domestic Humanoid Robots Unleashed This Year?

The domestic humanoid robot industry is booming, with major automotive companies entering the field.

What New Innovations Have Domestic Humanoid Robots Unleashed This Year?

The domestic humanoid robot industry is booming, with major automotive companies entering the field.

Tencent Opens Source Lightweight Video Generation Model HunyuanVideo 1.5, Deployable on Consumer-Level GPUs

Tencent opens sources an 8.3B parameter video generation model, deployable on consumer-level GPUs.

Tencent Opens Source Lightweight Video Generation Model HunyuanVideo 1.5, Deployable on Consumer-Level GPUs

Tencent opens sources an 8.3B parameter video generation model, deployable on consumer-level GPUs.

Meta Will Move React to the Linux Foundation

Meta will hand over React to the Linux Foundation for management

Meta Will Move React to the Linux Foundation

Meta will hand over React to the Linux Foundation for management

Kimi's K2 Open-Source Language Model Supports Dynamic Resource Availability and New Optimizer

Kimi K2 Open-Source Model Releases New Optimizer

Kimi's K2 Open-Source Language Model Supports Dynamic Resource Availability and New Optimizer

Kimi K2 Open-Source Model Releases New Optimizer

General

Packing Tickets! Fundamental AI Concepts Everyone in Frontend and Beginners Must Understand

A primer on AI fundamentals, covering symbolicism to deep learning.

Packing Tickets! Fundamental AI Concepts Everyone in Frontend and Beginners Must Understand

A primer on AI fundamentals, covering symbolicism to deep learning.

Announcing the Launch of ml-container-creator, Simplifying SageMaker BYOC

AWS releases a tool to simplify SageMaker BYOC deployment.

Announcing the Launch of ml-container-creator, Simplifying SageMaker BYOC

AWS releases a tool to simplify SageMaker BYOC deployment.

AI Training Camp Launches – E-commerce Professionals Must Learn: How OOTD Becomes a Bestseller

Alibaba Cloud launches OOTD e-commerce content generation tutorial based on Wan2.5

AI Training Camp Launches – E-commerce Professionals Must Learn: How OOTD Becomes a Bestseller

Alibaba Cloud launches OOTD e-commerce content generation tutorial based on Wan2.5

Unveiling the 'Theatrical Family' of JS Inheritance: Prototypes, Prototype Chains, and new

Explains JS prototype chain and inheritance mechanisms using a family analogy.

Unveiling the 'Theatrical Family' of JS Inheritance: Prototypes, Prototype Chains, and new

Explains JS prototype chain and inheritance mechanisms using a family analogy.

【Today's New Release】Pura X Compatibility Guide

Huawei Releases Pura X Foldable Screen Compatibility Guide

【Today's New Release】Pura X Compatibility Guide

Huawei Releases Pura X Foldable Screen Compatibility Guide

Finding Direction in Platform Maturity

Don't just focus on tools; the core of platform maturity is eliminating cognitive load.

Finding Direction in Platform Maturity

Don't just focus on tools; the core of platform maturity is eliminating cognitive load.

HarmonyOS Technical Q&A: Episode 10

HarmonyOS developer common technical questions answered in Episode 10

HarmonyOS Technical Q&A: Episode 10

HarmonyOS developer common technical questions answered in Episode 10

HarmonyOS Technical Q&A: Issue 11

HarmonyOS technical questions and answers series, addressing specific issues.

HarmonyOS Technical Q&A: Issue 11

HarmonyOS technical questions and answers series, addressing specific issues.

Primary

Amazon Web Services Introduces Remote Build Cache in ECR to Accelerate Docker Image Builds

InfoQ - Cloud•Published on November 21

Insight•Cloud Computing and Backend•Score: 9/10

ECRDockerCI/CDRemote CacheBuild Efficiency

Source

◆

From 'User Thinking' to 'Cultivation Thinking': Discussing the Arrogance and Bias of Tech Giants

Tech Scramble•Published on November 20

News•General News•Score: 8/10

User ExperienceTech Giant MentalityHarmonyOSXiaomiProduct Strategy

Source

◆

AWS Introduces Landing Zone Accelerator General Configuration and LZA Compliance Workbook

AWS - Security•Published on November 20

Insight•Cloud Computing and Backend•Score: 8/10

AWSLanding Zone Acceleratorsecurity baselinecompliancemulti-account architecture

📈

Perspectives

In my view, AWS's release of LZA general configuration is aimed at further solidifying its position in the enterprise cloud service market, especially in the face of increasingly complex compliance challenges. This tool saves customers money, which translates to earning their trust. In November 2025, cloud security and compliance are the top priorities for all enterprise IT departments. The emergence of such tools essentially transforms compliance from a 'manual task' into a 'configuration task', lowering the professional threshold and time cost for cloud adoption. Although domestic cloud vendors also have similar products, they still lag behind global giants like AWS in terms of international compliance standard coverage and toolchain maturity. In the future, 'out-of-the-box' compliance solutions will become a standard competitive advantage.

Source

◆

Using Application Load Balancer Target Optimizer to Improve Application Performance

AWS 网络•Published on November 20

Insight•Cloud Computing and Backend•Score: 8/10

AWS ALBTarget OptimizerConcurrency ControlLarge Model InferenceApplication Performance

📈

Perspectives

In my view, this feature is AWS's architectural optimization to better serve emerging computing paradigms like AI model inference. Previously, controlling backend concurrency with L7 load balancing was a pain point, and now it's providing a solution through the agent model. In November 2025, cost and efficiency of AI inference services are the main concerns. Target Optimizer, through fine-grained scheduling, can effectively maximize the potential of expensive resources like GPUs, reducing 'lag' and 'overload,' directly enhancing user experience and lowering operational costs. Domestic cloud vendors also face similar issues with their LB products, and I believe similar features will be introduced soon. After all, whoever can reduce computing costs the most will gain an advantage in the AI services market.

Source

◆

AWS re:Invent 2025 Amazon EKS and Kubernetes Conference Guide

AWS - Containers•Published on November 21

News•Cloud Computing and Backend•Score: 8/10

EKSre:Invent 2025Auto ModeAI WorkloadsPlatform Engineering

Source

◆

Meituan LongCat Interaction Team Releases Technical Report on Large Model Interaction System WOWService

Meituan Tech•Published on November 21

Insight•AI and Frontier Technologies•Score: 8/10

MeituanLarge ModelWOWServiceAgentInteraction System

Source

◆

No Data Upload Required! Apple Officially Open Sources Embedding Atlas, an Interactive Vector Visualization Tool for Desktop Using Rust+WebGPU

InfoQ - AI & LLM•Published on November 21

Insight•AI and Frontier Technologies•Score: 8/10

Embedding AtlasEmbedding VectorsVisualizationWebGPURust

Source

◆

Kiro is Now Available, Core Features Receive Enhancements

InfoQ - AI & LLM•Published on November 21

News•Software Architecture and Engineering Practices•Score: 8/10

KiroProperty-Based TestingCheckpoint ReversalAI Programming AssistantMulti-root Workspace

Source

◆

Achieving High Availability with Block-Level Replication in Amazon RDS for SQL Server Web Edition

AWS - Databases•Published on November 20

News•Cloud Computing and Backend•Score: 7/10

RDSSQL Server Web EditionBlock-Level ReplicationHigh AvailabilityMulti-AZ

Source

◆

Kiro 'Startup Support Program': AI Agent Helps You Scale from Idea to Production Without Worry

InfoQ - AI & LLM•Published on November 21

News•AI and Frontier Technologies•Score: 7/10

KiroAI AgentStartup SupportAWSSpec Mode

Source

◆

Finding Direction in Platform Maturity

Thoughtworks•Published on November 21

Insight•Platform Engineering•Score: 7/10

Platform MaturityDORA MetricsCognitive Load

Source

Essential

Top Strategies for Building Highly Available Systems

ByteByteGo•Published on November 20

Insight•Software Architecture and Engineering Practices•Score: 9/10

High AvailabilitySystem DesignResilienceFault ToleranceRedundancy

Source

◆

Linus Talks About AI and Atmosphere Programming: It's Good for Beginners, But Not for Production

InfoQ - AI & LLM•Published on November 20

News•Software Architecture and Engineering Practices•Score: 9/10

Linus TorvaldsAI programmingatmosphere programmingLinux kernel maintenanceRust migration

Source

◆

Introducing Strands Agent SOPs – AI Agents with Natural Language Workflows

AWS 开源•Published on November 20

Insight•AI and Frontier Technologies•Score: 9/10

Agent SOPsNatural Language WorkflowsAI AgentsStandardized ProcessesReproducibility

Source

◆

Exclusive Share from a Top-Tier CIO: How AI is Reshaping Developers' Future for the Next Decade

InfoQ - AI & LLM•Published on November 21

Insight•AI and Frontier Technologies•Score: 9/10

AI reshaping developmentdeveloper transformationAI efficiencyfull-stack engineerknowledge-driven AI

Source

◆

E215｜Crypto Miners with Electricity Advantage: Will They Be the 'Gold Miners' of the AI Era?

SV101 Podcast•Published on November 21

Insight•AI and Frontier Technologies•Score: 9/10

Crypto MinersAI Computing PowerElectricity InfrastructureFinancial RiskData Center

📈

Perspectives

In my view, this 'gold rush' in AI infrastructure is turning into a capital feast. Crypto miners entering the field are essentially shifting from low-efficiency 'virtual mining' to high-expectation 'AI gold digging', but their core capability remains securing cheap electricity and land. As of November 2025, many transitioners may just be swapping 'coin trading' for 'computing power trading'. High leverage and high-risk financial tools are being introduced, and the shadow of a bubble is growing. In the next two to three years (2026-2028), we may see many 'pseudo-gold diggers' being devoured by capital. The real opportunity lies with those who can stably provide green electricity and have efficient data center operation capabilities, not just those with a few old mining sites. Everyone is looking for electricity, and whoever can supply it long-term, stably, and legally will be the true 'water seller'.

Source

◆

Amazon S3 Introduces Attribute-Based Access Control for General Purpose Storage Buckets

AWS 新闻•Published on November 21

News•Cloud Computing and Backend•Score: 9/10

ABACS3IAMTag-Based ControlAccess Management

Source

◆

Dark Data Tax: How Stockpiling Poisons Your AI

Juejin Frontend Hottest•Published on November 19

News•Software Architecture and Engineering Practices•Score: 9/10

Data GovernanceData LakeAI Data QualityData ObesityCost Optimization

🔬

Tech Insights

The article introduces the concept of 'data obesity syndrome', pinpointing the issue not as low storage costs, but as low data value conversion efficiency. It states that 92% of total cost of ownership (TCO) is related to data conversion rather than storage. It introduces the Predator-Prey Framework to diagnose data ecosystem imbalance, treating data volume as 'prey' and data value (analysts, models) as 'predators', using the Lotka-Volterra equation to explain how dark data (unconsumed prey) leads to 'predator starvation'. This metaphor is very vivid and insightful. It proposes the Data Sustainability Index (DSI) as a quantitative metric, measuring data ecosystem health through 'total analysis computing time', 'total data lake cost', and 'proportion of active datasets', offering a very practical tool for measurement and management. It emphasizes the impact of signal density rather than dataset size on model performance, which is a revelation for AI teams, directly refuting the misconception that 'more data is always better', and highlighting the importance of data curation.

📈

Perspectives

In my view, the discussion around 'data obesity syndrome' and 'dark data tax' represents the inevitable shift in the data industry from wild growth to refined operations. Previously, everyone was racing to collect data, but now it's clear that collecting too much leads to a jungle of irrelevant data, which can't be turned into useful insights. In the domestic market, especially amid the AI large model craze, companies are feverishly collecting data to feed models, but few are actually paying attention to the 'nutritional value' of this data. This article serves as a wake-up call. Vendors that can provide data governance, data quality, data curation, and data value assessment solutions will have significant market opportunities in the coming years. It also means that data platform products will no longer be simple data storage and computation tools, but will need to integrate stronger capabilities in data lineage, data cataloging, data quality monitoring, and usage analysis, helping enterprises achieve true 'data slimming' and 'value building'.

Source

◆

What New Innovations Have Domestic Humanoid Robots Unleashed This Year?

InfoQ - AI & LLM•Published on November 21

Insight•AI and Frontier Technologies•Score: 9/10

Humanoid RobotsEmbodied IntelligenceLarge ModelsCommercializationChinese Market

🔬

Tech Insights

Xpeng's IRON features a 'Physical World Large Model' and self-developed Turing chips, with a computing power of 2250 TOPS, indicating that automotive companies are developing full-stack self-research capabilities in AI chips and embodied intelligence, no longer just acting as simple hardware integrators. Yushu's H2 adopts an 'Efficient Balance' design philosophy, which in the context of commercialization for general-purpose applications, ensures both smooth motion and system reliability—this is the key to real-world scenario deployment. Galaxy General's Galbot features end-to-end embodied intelligence large models GraspVLA and GroceryVLA, directly solving the generalization grasping challenge in retail scenarios, marking a significant step from the lab to real-world applications. Embodied intelligence's scenario-specific breakthroughs hold more commercial value than general capabilities alone. Zhiyuan's G2 has been deployed in industrial scenarios, particularly in automotive parts production and logistics sorting, proving that humanoid robots are beginning to demonstrate potential to replace human labor in specific industrial processes.

⚠️

Notable

Although the article lists many advancements, it does not elaborate much on the technical barriers and cost challenges behind these 'new moves'. For example, the cost and mass production difficulty of high-degree-of-freedom dexterous hands, as well as the data sources and generalization capabilities of the 'Physical World Large Model', are the real concerns we have—the 'hard nuts to crack'. Major companies are rushing in with huge investments, but many are still at the stage of lab prototypes or small-scale pilots. Huang Renxun of NVIDIA's optimism contrasts sharply with LeCun of Meta's warning that 'we don't even know how to make robots smart enough to be truly useful', indicating the industry is still in the early bubble phase, with long commercial payback cycles. Additionally, while scenarios like 'human-car-home', robot tour guides, and elderly assistance are appealing, the article barely mentions user acceptance and ethical safety issues, which are critical hidden costs before widespread adoption of embodied intelligence.

📈

Perspectives

In my view, the domestic humanoid robot market is now the 'Everything Can Be AI' scenario. Capital and giants are flocking in, but only those that can solve specific industry pain points, achieve cost control, and realize scalable deployment will emerge successfully. The entry of internet/ecommerce giants like ByteDance, Alibaba, JD.com, Baidu, Tencent, and Xiaomi, as well as automotive companies such as Xpeng, Chang'an, SAIC, and BYD, essentially signifies a competition for the future smart ecosystem's entry points and platform control. Whoever can establish a complete closed-loop from the 'brain' (large model) to the 'body' (robotic platform) and then to the 'application scenarios' will gain pricing power in the next decade. Currently, industrial manufacturing and logistics sorting are likely to be the first large-scale deployment scenarios, as they offer clear return models and high demands for precision and efficiency. The consumer market, such as home services, remains concept-driven in the short term, and will only become practical if costs can be reduced to the level of 10,000 yuan. Huang Renxun's five-year popularization theory seems overly optimistic to me, but the trend is irreversible.

Source

◆

Tencent Opens Source Lightweight Video Generation Model HunyuanVideo 1.5, Deployable on Consumer-Level GPUs

InfoQ - AI & LLM•Published on November 21

News•AI and Frontier Technologies•Score: 9/10

HunyuanVideo 1.5DiT ArchitectureText-to-VideoConsumer-Level GPUSSTA Sparse Attention

Source

◆

Meta Will Move React to the Linux Foundation

InfoQ - Frontend•Published on November 20

News•Frontend Development•Score: 8/10

React FoundationLinux FoundationMetaVercelTechnical Governance

Source

◆

Kimi's K2 Open-Source Language Model Supports Dynamic Resource Availability and New Optimizer

InfoQ - AI & LLM•Published on November 21

Insight•AI and Frontier Technologies•Score: 8/10

Kimi K2Mixed Expert ModelMuonClipOptimizerOpen-Source Model

Source

General

Packing Tickets! Fundamental AI Concepts Everyone in Frontend and Beginners Must Understand

Juejin Frontend Hottest•Published on November 18

Insight•AI and Frontier Technologies•Score: 7/10

Artificial IntelligenceMachine LearningDeep LearningNeural NetworkFundamental Concepts

🔬

Tech Insights

The article's strength lies in its extreme simplification, explaining complex concepts like 'weights', 'activation functions', and even 'AI winter' through relatable analogies such as 'training a dog', 'building with LEGO', and 'layering a hamburger'. This truly achieves the goal of 'everyone can understand'. For non-technical product managers or those looking to quickly transition into FE/BE roles, this is an excellent starting point. The 'apple recognition' example in connectionism uses simple weighted sums and introduces non-linear functions through the XOR problem, clearly structured logic chains, which are more intuitive than many academic textbooks. This reminds me of when I was mentoring new hires, struggling to find suitable introductory materials. This article is a model for hands-on understanding of core principles.

⚠️

Notable

As a 'popular science' article, it deliberately sacrifices depth. For architects or senior engineers, this article is only suitable as a dessert after a meal, offering little real value. The analogies like 'linear function' and 'two lines' solving the XOR problem, while easy to grasp, highly abstract the technical details and underlying principles, which may lead some beginners to mistakenly believe that AI is 'not that complicated'. The article's content, although published in November 2025, discusses classic concepts from the early development of AI. In my opinion, AI is evolving rapidly now, especially in areas like agents, multimodal systems, and fine-tuning large models. This article does not mention any of these frontier advancements, making it feel slightly outdated. If a team truly wants to understand AI deeply, this is only the first step; there is still a lot more hard-core content to digest.

📈

Perspectives

In 2025, the emergence of such 'AI popular science' articles on the market itself highlights a problem: AI education in China is still far from widespread. Many companies and individuals claiming to 'embrace AI' actually lack a basic understanding of its concepts. The appearance of this article reflects a strong demand in the market for 'fundamental, easy-to-understand, and non-intimidating' AI introductory materials. For our technical teams, this means that, at the current time (November 2025), if we are to push forward with AI transformation or AI empowerment, internal cognitive alignment and foundational training should be the top priority. Don't expect people to actively read obscure academic papers; this kind of 'popular science' article is actually a key stepping stone for promoting AI concepts. However, we should also be cautious about this overly simplified approach, which may lead people to develop unrealistic expectations about AI's capabilities and implementation costs. In the end, it's important to return to practical business scenarios and technical stacks, and not treat AI as a universal solution.

Source

◆

Announcing the Launch of ml-container-creator, Simplifying SageMaker BYOC

AWS 开源•Published on November 20

Insight•AI and Frontier Technologies•Score: 7/10

SageMakerBYOCMLOpsContainerizationAWS

Source

◆

AI Training Camp Launches – E-commerce Professionals Must Learn: How OOTD Becomes a Bestseller

InfoQ - AI & LLM•Published on November 20

Insight•AI and Frontier Technologies•Score: 6/10

OOTD GenerationWan2.5 of Tongyi WanxiangAlibaba Cloud BailianAI E-commerce ApplicationsZero Foundation AI Design

Source

◆

Unveiling the 'Theatrical Family' of JS Inheritance: Prototypes, Prototype Chains, and new

Juejin Frontend Hottest•Published on November 20

Insight•Frontend Development•Score: 6/10

JavaScript PrototypePrototype Chainnew Keyword__proto__Inheritance Mechanism

Source

◆

【Today's New Release】Pura X Compatibility Guide

InfoQ - Frontend•Published on November 21

News•Frontend Development•Score: 6/10

Pura XFoldable Screen CompatibilityHarmonyOSExternal Screen DevelopmentUI Compatibility

Source

◆

Finding Direction in Platform Maturity

Thoughtworks•Published on November 21

Insight•Platform Engineering•Score: 6/10

Platform MaturityDORA MetricsCognitive Load

Source

◆

HarmonyOS Technical Q&A: Episode 10

InfoQ - Frontend•Published on November 21

Insight•Frontend Development•Score: 5/10

HarmonyOSScroll componentEventHubImage croppingCity location

Source

◆

HarmonyOS Technical Q&A: Issue 11

InfoQ - Frontend•Published on November 21

Insight•Frontend Development•Score: 4/10

HarmonyOSDeveloper Q&AUI AdaptationComponent UpdatePicture-in-Picture