W21 - Cross-Platform and Containers

I'm preparing the next session of the fundamentals mentorship series and, after some thought, I want to talk about cross-platform technologies and containers. Although this topic is large and complex, I hope, like the previous YuLu session, it can provide some useful and correct insights.

Why do I consider cross-platform skills to be fundamental?

1. Reducing development costs and improving deployment efficiency are the driving forces behind cross-platform technology. With the promotion of HarmonyOS, this trend will only strengthen. Take common mainstream platforms as an example: you need to support mobile (iOS, Android, Harmony), desktop (macOS, Windows, Linux), and the web. While native technologies can deliver the best user experience, relying solely on native development for such multi-platform demands is clearly inefficient and costly.

2. There are many proprietary cross-platform technologies within the company, and in daily work we inevitably encounter container-related issues like Titans, MSC, MRN, and others. The financial services frontend has also evolved from EHC to Neo to Recce. A systematic understanding of cross-platform technology helps pinpoint the root cause of problems and improves efficiency. In the era of AI-assisted coding, the need for systematic understanding is even more pronounced. Timely and effective troubleshooting, as well as application security and reliability, are not things AI excels at; they require people to apply systematic understanding and judge solutions in the context of the actual processes.

3. Looked at within a broader technical context and a longer historical perspective, cross-platform development has been continuously evolving since the birth of computing. Each mainstream solution did not arise in isolation but is part of an interconnected web. Only by recognizing this web can we understand the uniqueness of each technology and make accurate choices and judgments. Early in my career I wrote many cross-end applications—React Native, mini programs, and so on. Back then we used whatever the team had, mainly to learn a craft. I had no idea where those technologies came from, and no matter how much I wrote, I couldn't form high-quality judgments.

This time I also want to explore collaborating with AI to produce content: from the initial content organization, to compressing the manuscript text, to quickly generating ByteByteGo-style system diagrams, and even trying to include a short podcast segment.