turso In-memory mode

Add support for in-memory mode like SQLite has because it's great for just playing around with a database.

Jun 20 '24 06:06 penberg

hey @penberg . I would love to work on this task.

Jul 09 '24 20:07 jatin510

Hey @jatin510, feel free, although I think it's more rewarding to do this when we support writes (#45) because then you can actually do something in-memory.

Jul 11 '24 13:07 penberg

Sure, I will start working on this issue. Once we will start support writes.

Jul 11 '24 13:07 jatin510

~I am not sure, maybe it could implement live read-only access to the existing sqlite3 instances via Shared Cache mode. https://www.sqlite.org/sharedcache.html~. Seems like it is only in-process cross-thread infra.

Jul 21 '24 19:07 sitano

There is now experimental write path as of 01ec5f5e6472e0b7f752e1a3e47e560a19f34c42 so now an in-memory I/O is something that makes more sense. Probably needs CREATE TABLE support (#146) first to be useful.

Aug 01 '24 16:08 penberg

We have in-memory mode. It could be improved, but that deserves a new issue.

Jun 26 '25 13:06 penberg