英文写作习作。
This week, I learned a few things I hadn’t thought about before from an event called the “First Local-First Meetup” in China.
On September 20, Zixuan Chen gave a talk titled “CRDTs Are Not Enough: 从 CRDT 到本地优先同步引擎.”
What interested me most, however, was not CRDT (Conflict-free Replicated Data Type, 无冲突复制数据类型) itself, but a more fundamental question:
What exactly does “local-first” mean? And how is it different from “local-only”?
Before this week, my understanding of local-first was quite simple: data should stay local whenever possible, and cloud services should only play a secondary role.
At first glance, this sounds very similar to “local-only.” But after thinking about it more carefully, I realized that they describe two quite different ideas.
Local-only is really easy to understand. Your data stays on your own device, or within infrastructure that you control. The application does not depend on a remote server to store, synchronize, or process that data.
This gives the user a very clear boundary: if the local system still works, the application still works.
Local-first is different.
It does not necessarily reject servers or the cloud. Instead, it treats the local copy of the data as the primary copy from the user’s point of view.
The application should remain useful without a network connection, changes should be written locally first, and synchronization should happen later when other devices become available.
So, to me, the key difference is that local-only tries to avoid dependence on remote services, while local-first tries to make local data the center of the user experience.
I think local-first is a more practical idea for modern applications. It keeps many advantages of local software, while still allowing synchronization and collaboration when needed.
留言 / CONVERSATION
欢迎交流,评论将在审核后展示,你的邮箱不会被公开。
正在准备评论区…