Pekka Enberg
Pekka Enberg
@LtdJorge @PThorpe92 Is there some other name that you think is better? I think "syscall-based I/O" is the right technical term here, but if it means nothing to people, then...
@rajivharlalka You can test the function with a modifier: ``` sqlite> SELECT unixepoch('2024-08-05'); 1722816000 ``` Of course, you won't be able to cover all of the functionality in `unixepoch()` but...
Hey @rajivharlalka! Are you still working on this?
Hey @rajivharlalka, would you be able to move the tests to the new file as pointed out by @sonhmai?
@jatin510 Nice! We want to do writes to a WAL so persistence can come later when we have it.
The `get_next_available_page()` thing won't work so at least that ought to be dropped. If you look at what SQLite does for `CREATE TABLE`: ``` sqlite> EXPLAIN CREATE TABLE users (id...
@jatin510 There's now work on write path: https://github.com/penberg/limbo/pull/229 When that lands into `main`, we can resume work on schema management.
We now have write path in 01ec5f5e6472e0b7f752e1a3e47e560a19f34c42 so work on `CREATE TABLE` can now be implemented!
It's good to study what SQLite does and check alternatives like https://duckdb.org/2021/08/27/external-sorting.html.
No, that's not how it works exactly. There is a `SorterOpen` instruction that opens a transient index that a VDBE program uses for sorting. The `OpenPseudo` instruction, on the other...