We use VARCHAR for almost everyone, and NVARCHAR is very rare.
NVarchar does not need the product code - we do not allow anything in them except AZ, 0-9 and "_" ...
Two times more storage space, but also has only half the entries on the index page (and data page), and half the memory cache is wasted, there are more CPU cycles to compare data, etc.
IME commonly used foreign accents only work in Varchar (i.e. LATIN-1). We have no plans to make Chinese or other alternative character sets, and when we can deal with this character set using NVarchar from the first day, we will be least worried - text alignment from right to left or vertical ?: (
And if you enable NVarchar, say, a name, how are you going to enter the extended charcater from the keyboard? And if you import data (as it is already NVarchar), how can you search for this client using the standard QWERTY keyboard. There are many, many things related to the internationalization of the application, so I believe that it makes no sense to "resolve this using NVarchar."
But again, there are many places where I go to have NVarchar ... and most columns also have a width of 50 characters ... they should know something about population growth plans and expanding zip code plans that I don't know!
Kristen
source share