On 05/30/12 10:17 AM, Khangelani Gama wrote:
> So talking about compatibility, you are saying we can continue using
> UTF-8?, but this will create more work for us because most of our
> scripts assume that encoding is SQL_ASCII hence we want continue
> using SQL_ASCII in Postgres 9.1.2.
SQL_ASCII is not really an encoding, its saying "this data has no
encoding at all, its just bytes". UTF-8 will reject any data thats not
properly UTF8 encoded.
converting from 'undefined' encoding to a rigorously enforced encoding
is problematic. On the other hand, working in SQL_ASCII has all kinda
ugly issues, like length(somestring) is just counting bytes, and not
characters if the string happens to be a multibyte encoded entity.
collation order is just binary. upper/lower don't work on anything
other than USASCII (eg, accented characters are ignored).
sounds to me like you're stuck in SQL_ASCII
--
john r pierce N 37, W 122
santa cruz ca mid-left coast