Re: Oracle 8 NLS Characterset

Hi --

jimg_ppg_at_my-dejanews.com wrote:
>
> I am in the process of developing an Oracle 8 database on an HP UNIX server.
> I would like to store text in the database in multiple languages which
> include, but is not limited to, Japanese, Korean, Chinese, French, Spanish
> and of course American English. My question(s) is this: Is there a single NLS
> Characterset value that I can use that will accommodate all these languages
> characters and accent marks? If so, what is the setting? If not, how many
> database instances must I create to accommodate all and what settings will
> accommodate what language?

Well there are two cases:

• Common part for the two case:
  • install Oracle using the character encoding AL24UTFFSS for Ora7.3.4 or even better UTF8 for oracle 8.0. (see National Language Support)
• Your client must able to manage several languages:

Then the client must define, in its environment, the variable NLS_LANG to <whatever>_<whatever>.AL24UTFFSS
For example: setenv NLS_LANG 'AMERICAN_AMERICA.AL24UTFFSS'

And you have to do the translation between the client and oracle by hand.

• Your client see only one of the languages

Then the client define the same variable to the correct value and oracle will translate, on the fly, from AL24UTFFSS to the given encoding. For example: setenv NSL_LANG 'FRENCH_fr.WE8ISO8859P1'

Our application is using Oracle 7.3.4 on Solaris and is able to manage all the language that you are talking about. We have to do the translation on our own.

See National Language Support for all the possible values.

Hope it helps.

        Christophe

	Christophe L. Galerne                      cgalerne_at_tumbleweed.com
	Software Engineer                             Tel : (650) 569 3683
	Tumbleweed Software Corp.                http://www.tumbleweed.com

	2010 Broadway Street                        Redwood City, CA 94063
	"La générosité légendaire des Corses se vérifie une fois encore, 
	 malgré les tabous attachés au don du sperme"          Corse-Matin