Oracle 12c Character Set - Accents (Tildes)

Hi,

We are having an issue with accented words on a RHEL7 Linux VM with Oracle 12c.

The current characterset is AL32UTF8, and I always though it would be able read words with accents (i.e. término), but it's not as I expected.

For example,

SYS_DEV2>insert into test (name) values ('término');

1 row created.

Elapsed: 00:00:00.00

SYS_DEV2>select * from test;

NAME

--------------------------------------------------

t??rmino

And when I set the NLS_LANG environmental variable to the same character set, it can now read those type of words:

export NLS_LANG=AMERICAN_AMERICA.AL32UTF8

SYS_DEV2>insert into test (name) values ('término');