Gauthier

would you know which `execute` call in `JDBC3Statement` is throwing ? I don't know which version was used for the stacktrace provided, so difficult to know which line we are...

> JDBC3Statement My bad, i didn't read properly. So the not implemented method is this one: https://docs.oracle.com/javase/7/docs/api/java/sql/Statement.html#execute(java.lang.String,%20int) Would you be able to find out what is the second parameter passed?...

`supportsGetGeneratedKeys` returns `true`, but that is not implemented everywhere from what i can see. I think it's because the support for that in SQLite is limited via `last_insert_rowid`. At least...

> Maybe `supportsGetGeneratedKeys` should not return true if that support does not actually exist in the driver. It does, but not everywhere. That's my understanding.

> Hm... would there be any considerable downside to letting `public boolean execute(String sql, int autokeys)` call `public boolean execute(String sql)`, silently ignoring the `int autokeys`? That would be a...

> I consider having that in the Statement as nice-to-have, but not essential. I'm fine with executing `SELECT last_insert_rowid();` when necessary. You can use the JDBC way via getGeneratedKeys, it...

What kind of warnings would you expect ? I am not familiar with JDBC warnings, but i suppose they would carry some information coming from the underlying database.

> i.e. any warnings that appear in other sqlite clients but are unavailable using this library (say, after a fk constraint fails) Maybe provide some context or examples, as you...

You're not really helping your case here with this attitude.

How would that be made available in the driver though ? Would this be through the JDBC warnings (#483) ?