NLS_LANG and Unicode

From: Matt <mccmx_at_hotmail.com>
Date: 17 Sep 2003 00:51:36 -0700
Message-ID: <cfee5bcf.0309162351.2452f34e@posting.google.com>

Hello all,

After a recent migration of an 8.1.7 database from Western European P15 to UTF8 character set, I am concerned we may have an incorrect setting for client side NLS_LANG variable.

My understanding is that NLS_LANG should always match (correspond to) the operating system code page/character set.

Following this logic suggests that it should be set to WE8MSWIN1252 on my WIN 2K middle tier server. It is currently set to WE8ISO8859P1 which is apparantly equivalent to MSWIN1252.

However the application (Peoplesoft HR) supports a variety of clients (browsers) who use a variety of languages (Polish/Turkish, etc). Since Peoplesoft is supposed to be fully Unicode compliant, Peoplesoft recommend setting the NLS_LANG to UTF8 on the client(i.e. middle tier).

Therefore I need to change this setting on my middle tier server but I am concerned about the implications.

If I change the NLS_LANG setting on my middle tier server from WE8ISO8859P1 to UTF8, will the current data become corrupt....?

Thanks very much

Matt Received on Wed Sep 17 2003 - 02:51:36 CDT