Better than Json

[iceBear67]

显然，Json 并不适合大量机器和物品数据的序列化。相对比其他的二进制文件序列化方案（例如 kyro ），他实在是太慢了。

我们应当将 Json 更换为更快的实现，并且考虑引入数据版本机制来制作多个版本之间的兼容。不过也可能没有这个必要

[iceBear67]

不过我们同时需要一个能够审查这些数据的工具。或许我们需要新开一个 AstralFlow Developer Tools? Kyro 看起来也不是很符合现在的设计。

[iceBear67]

We may consider Jackson OR DSLJson as a optional opinion for large-scale networks.

[iceBear67]

Data-versioning is still considerable. We may introduce it before releasing

[iceBear67]

Benchmarks needed.

[iceBear67]

This isn't related to any breaking change on user-level, which allows us to put this issue lower priority.

And benchmarks needed.