Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> It's interesting that you want to go in 100% the opposite direction
> from Kevin, who seems to want to eliminate type inference
> altogether. Maybe our current compromise isn't too bad, if it makes
> everybody unhappy in opposite directions ;-)
Well, it's probably worth noting that, while I would prefer strong
typing all around, I recognize that unless that's easier than I think
to make as a configuration option, and strict is not the default, it's
not going to happen. Too many people prefer things the other way.
What I'm most concerned about are the corner cases where strict typing
would give one non-error result and the inferred typing results in an
error or a different result from the strict typing. I'm willing to
argue that those are bugs, at least when the strongly typed behavior
is mandated by the SQL standard.
I pray that I never again have to deal with a database as generous
with typing as Pick or Advanced Revelation. Seriously, working with
those gave me that feeling you normally only get during a nightmare.
:-( When a "date of birth" column can contain a list of phone numbers
-- well, you want just want do something that makes you forget you
saw that....
-Kevin