Problems with character_sets/collations when importing data

[john d. herron]

Just created a database and a table to test character set and collation handling.
Working with mixed French- and Italian-language (ISO 8859-1 or cp1252) data.

Linguistically, the database is configured thus:
+--------------------------+----------------------------+
| Variable_name | Value |
+--------------------------+----------------------------+
| character_set_client | latin1 |
| character_set_connection | latin1 |
| character_set_database | latin1 |
| character_set_filesystem | binary |
| character_set_results | latin1 |
| character_set_server | latin1 |
| character_set_system | utf8 |
| character_sets_dir | /usr/share/mysql/charsets/ |
| collation_connection | latin1_swedish_ci |
| collation_database | latin1_swedish_ci |
| collation_server | latin1_swedish_ci |
+--------------------------+----------------------------+

Successfully tried SET NAMES to change collation_connection and collation_database to latin1_general_cs; but attempting to do the same for collation_server (in my.cnf) has so far not worked.
Looks like adding

# * Character-set settings
#
collation_server=latin1_general_cs
character_set_server=latin1

in the [mysqld] section is not catching on.

Importing a latin1 (cp4...) csv test file via LOAD DATA INFILE resulted in a table where all of the accented characters are mangled.

Could the affected three fields of the table perhaps be 'saved' by applying some SELECT CONVERT magic?

The server is a Debian (Ubuntu) 5.0.51 vintage.

Any help will be appreciated.