Zach Musgrave

My PR from months ago claims to fix this but I think that's a lie. I think it's still possible to pass in more bind vars than the query contains,...

I think we need a simple repro for this. Summarizing what I understand: 1) Table has no null constraint. 2) Make a commit that deletes a row that has a...

First milestone is released: reads of dolt_branches get their data from the database's state at transaction start time. Writes are still not safe, can concurrent calls to `dolt_branch` can fail...

#benchmark

`@@max_connections` returns the correct value but changing its value on a running server doesn't change the behavior. You have to `set @@persist.max_connections` and then restart the server for changes to...

Hi @Syuparn, There aren't specific plans to create an executable main package, but we would certainly take a contribution that does this to start a local server with the in-memory...

The above isn't true: the indexes do get created, but they don't speed up query execution.

Yes, unfortunately the in-memory database is not completely threadsafe right now. We have this on our roadmap to address, but it's a bit of work to accomplish.

Hi @jonahzheng, There is no persistent database implementation built into the package. The provided `memory` implementation only persists in memory for a single process's lifetime. You would need to implement...

This is now technically feasible with the addition of `schema` elements to serialized table names.