Hi Jeff,

Nice code! Clever how you did that. I knew that it could be done in t-sql (although I don't think I would have come up with a function as clever as yours). I like C# for things like random numbers and complex string manipulation. The functionality built in is more elegant IMO. The simple update statement you listed is great too, but it looses the ability to specify a lower and upper range to the random numbers. Like you said, performance is pretty equal. The CLR outperforms on very big data sets (1million+), but only by a small amount .

Anyone who reads this post gets to choose the option best suited to their particular situation.