Hi Peter,

Quote Originally Posted by Peter van Mil View Post
The collating sequences starting with "SQL_" are used for compatibility and it is a "feature" that they ignore the "-" in indexes. So probably Latin1_General_CI_AS works better (if you don't want this).

And that would be fine if the DataFlex string comparison collating rules matched. Unfortunately, you can't have the database engine returning valid records for DataFlex to then invalid the selection because the string comparison are based on some other collating sequence. Seriously bad news.

Quote Originally Posted by Peter van Mil View Post
I am going to test with SQL Server 2019 and Latin1_General_100_CI_AS_SC_UTF8. Until now it works, but I have done very little with Unicode values.
Peter, I think that is an excellent idea and in particular with the DataFlex 2020. Look forward to reading what you find.