We are using SAP Sybase SQL Anywhere 16 with C#.NET and want to use unicode characters in string fields. When setting up the database with UCA collation, this already works with VARCHAR fields. Unfortunately we had to learn that the max length is specified in bytes and not in characters, which makes it unusable for us.
What's the recommended solution?
- use
NVARCHAR(n)(UTF-8) - use
VARCHAR(n CHAR)(character-length semantics) - something else?
What's the difference/advantages between the 2 solutions?
Does the recommendation depend on the typical use and if there are many or few special characters? We have japanese customers where nearly all characters will be multi-byte, but we also have english ones.
We could also have unicode characters in key fields used for joins. So we plan to change the types for ALL string fields. Will there be a big performance loss after this change?
Any other suggestions?
Nice to have: additional interesting information would be how this all differs from MS SQL Server...
