Posted by Ed Murphy on 12/24/07 20:33

jefftyzzer wrote:

> I suppose that when you then join the fact to the time dimension, you
> could apply a COALESCE to the date column, e.g.,
> COALESCE(time_dim.date, 'Date not Available').

Don't you have to CAST or CONVERT time_dim.date to CHAR or VARCHAR
first, lest it try and fail to convert 'Date not Available' to
DATETIME instead?

[Back to original message]