Tom Lane wrote:
>Personally I don't think that this is a "transitional issue" and we will
>someday all be happy in upper-case-only-land. Upper-case-only sucks,
>by every known measure of readability, and I don't want to have to put up
>with a database that forces that 1960s-vintage-hardware mindset on me.
I think the SQL standard is screwy here on at least two levels.
Not only is upper case fuggly (we all seem to agree on that
point), but I think case folding is a Bad Idea in general. I
think the only time you should have to quote a DB identifier is
when it conflicts with a reserved word. Either be case sensative
or don't. I'm all for the (ignore but preserve case) way of doing things.
But it IS the standard, and as such, as much as we all seem to
dislike it, I believe it is better to follow it. You can't just
go around picking and choosing the standards you'll adhere to.
Like Microsoft. If it bothers you that much, put some effort
into changing it. Attain world domination and then force the
world to bend to The Right Way. Get rich and pay off enough
members of the standards body to get it changed. But until then,
live with it.
Now, I am all for configurability, and lots of it. By all means,
allow us to choose how we'd like case folding to be carried out,
or whether case folding (blech) is done at all. While you're at
it, allow us to choose whether NULL is treated as
zero/blank/empty or as SQL standard NULL. Allow us to force the
DB to do case-insensative comparisons on all character data.
Allow us, as DB admins, to f*** with the standard behavior until
we have a working mimic of MySQL or MS-SQL :-)
But I think the default behavior should adhere to the SQL
standard as closely as possible, even when we all hate it with a passion.
Just my $.02
Glen Parker
