Michal Sniatala

Yes, for write-type queries we should have a boolean result.

Thank you for looking into this. I guess we should check if other drivers behave the same way and fix it if necessary.

I guess the MySQLi driver was implemented/ported as first, so that's why we have it implemented this way. I will try to look into PostgreSQL. The only thing I'm concerned...

> Should I create a separate issue for that? It seems problematic to me when we are creating ResultInterface objects when resultID is a boolean. Yes, please. We will look...

Yes, that would work better.

You know... I am not convinced of any version (even my own). Maybe kenjis will have an opinion on this - we can always stay with what we have, too.

> FYI, if we allow setting the name manually this won't apply to SQLite. SQLite does not really support naming the constraint. Yes, I'm aware of this. Since we had...