On Sun, Nov 18, 2012 at 1:29 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> David Johnston <polobo@yahoo.com> writes:
>> On Nov 17, 2012, at 20:43, John R Pierce <pierce@hogranch.com> wrote:
>>> I think he meant a tiny tiny bit faster, primarily due to not having to validate the length.
>
>> Maybe... But I would presume a "varchar with no limit" does not validate length...
>
> There is overhead from the type system for varchar, whether or not it
> has a length limit --- you'll get at least some RelabelType nodes in
> expression trees, and those don't have zero cost to execute.
>
> I'd generally recommend using "text" if you don't have any interest in
> enforcing a specific length limit.
>
> regards, tom lane
>
>
> --
> Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-general
As far as I know varchar(n) with n being less or equal than 126 cannot be
toasted, and have only one octet of overhead.
http://www.postgresql.org/docs/9.2/static/storage-toast.html