A new name for storage engine?

[skyzh]

The current on-disk storage engine is called secondary. That's a long story to be told...

As @MingjiHan99 bootstrapped the RisingLight project, we expect to implement a SQLite-like on-disk storage engine -- single file, update in-place, textbook concurrency control. But later I thought the old-school storage engine wasn't a good fit for a modern educational database. At that time, I came up with a new design, the merge-tree column store engine. Everything on disk is immutable, which greatly simplified the concurrency control in our system.

Months ago, there were two folders in storage: disk, memory. How should we call the second storage engine? It turned out to be secondary, and we planned to develop all of the three engines, and make a shared trait for them to support adaptive storage engine for RisingLight.

Later, we decided not to continue with the disk engine development. So now we only have two engines -- memory and secondary.

And I believe now it's a good time to determine a new name for secondary. If you have any idea, feel free to post in this discussion.

[xxchan]

I finally know why it is called secondary 🤣🤣🤣