Ok, so something I need to do is a bit of investigation into how collations are managed, and how you could use .NET code to compare a string in the same way that any given collation would...
So, for this, I need to try and create a data set which contains all possible two-character unicode strings. Seeing as this will end up being 4 billion rows, it could potentially be quite (very) time consuming.
So I thought I would ask you guys how you would go about doing it.
A bit of background:
This is because I need to be able to take the output of two ordered SELECT statements from two DBs with identical schema, and in .NET be able to tell which row would come 'later'. This is so that if the databases contain different data, I can read from the side which has extra rows until the streams are back in sync. This is pretty easy for numeric types, but I need to work with string types too.