We need to define a column that will contain externally provided identifiers. Their
(1) length will vary from seven to ten characters
(2) all English letters
(3) both upper and lower case mixed with numeric digits
(4) with an even distribution of lengths <--missed this point when answering, thus got it wrong [I said varchar(10)]
Which of these data types will be more efficient?
Correct answer nvarchar(25)...
(1) I don't care how many people you asked in management, this will change next week as soon as this is in PROD...you know that's true
(2) for now, but when you least suspect it/expect it you will have worry about localization...
(3) anh, moot point...see point 2
(4) unless this is a lookup table, this can only be the case if the number of records is evenly divisible by four, as you add one record at a time...unless all your batch inserts are also divisible by four and only then do you insert.
good question though, since the parameters were set as hypotheticals, but in the 'real' world those first set of constraints ALWAYS are wrong...LOL