On Sat, Apr 26, 2008 at 08:33:28PM -0400, Robert Treat wrote: > enum types custom ordering. It also showcases the idea of data definitions > that "should never change", but that do changes every half dozen years or so. > Now you can argue that since it is expected that the ratings might change in > some way every few years that an enum type is not a good choice for this, but > I feel like some type of counter-argument is that this is probably longer > than one would expect thier database software to last. :-)
I think that if you are building software on the premise that it's only going to last five years, you oughta have a look around on the Internet again. Or think about why banks spent the money they did a few years back poring over ancient code making sure that two-digit year representations weren't in use. You can _of course_ make this sort of trade-off: the cost of the upgrade might be worth the natural ordering and boost in performance. But that was part of my point when noting that enums oughta come with a warning (and why I compared them to char()). I'm not arguing that they're completely useless; just that, like any oddly specialised tool, they require careful use. A -- Andrew Sullivan [EMAIL PROTECTED] +1 503 667 4564 x104 http://www.commandprompt.com/ -- Sent via pgsql-general mailing list (pgsql-general@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
