some unicode problem

I am facing a problem here I just can't figure out.

The database is 11.1.0.7, running on Red Hat Linux 4 and the character set is WE8ISO8859P1.

When running the query:

select col1 from tab1 where col2 like '%zoran%'

I get 'kort/zoran-ðinðiæ' as an row, but when running

select * from tab1 where col1 = 'kort/zoran-ðinðiæ'

I get no rows back.

Both col1 and col2 is nvarchar2

Any suggestions?

/johan

--
http://www.freelists.org/webpage/oracle-l