On Wed, Nov 24, 2010 at 4:43 AM, Thomas Kellerer <spam_eater@xxxxxxx> wrote:
Grzegorz JaÅkiewicz, 24.11.2010 10:37:I always try to convince people of this as well, but when they ask me under which circumstances this could happen, I can't think of a proper example.
just never use SELECT *, but always call columns by names. You'll
avoid having to depend on the order of columns, which is never
guaranteed, even if the table on disk is one order, the return columns
could be in some other.
Does anybody have an example that would show this?
Regards
Thomas
Even if an example doesn't exist, you can at least imagine a scenario where an improvement to postgresql is made such that the column order is decided internally, rather than by table definition. If the warning isn't given now, that improvement won't be possible. So I read that as "Don't rely on the table order, it's not part of the interface/contract and we're going to change it if we want to".
Derrick
