"Peter J. Schoenster" <peter(at)schoenster(dot)com> writes:
> On 22 Mar 2001, at 10:05, Tom Lane wrote:
>> There is *no* performance advantage of CHAR(n) over VARCHAR(n).
> I wonder if this question of char/varchar is postgresql specific or
> rdbms in general.
It's definitely RDBMS-specific. My comment applied to Postgres, which
stores CHAR(n) and VARCHAR(n) in essentially the same fashion --- it
doesn't really exploit the fact that CHAR(n) is fixed-size. (Mainly
because it's *not* fixed size in PG, what with TOAST, multibyte, etc.)
On other DBMSes there could be a difference, especially if the DBMS has
performance problems with variable-length fields.
regards, tom lane