Re: Advantages - if any of - of us ascii7 charset over ascii western european

"Paul Brewer" <paul_at_paul.brewers.org.uk> wrote in message news:3ddc2879_3_at_mk-nntp-1.news.uk.worldonline.com...
>
> "Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message
> news:1_vC9.79994$g9.225378_at_newsfeeds.bigpond.com...
> >
> >
> > Strongly recommended, I would say!! (Bear in mind that UTF8 or
UTF16 -with
> > various AL codes prefixed to them- are compulsory for 9i databases as
> their
> > national character sets).
> >
> > Regards
> > HJR
> >
> Good point, Howard.
> In 8.1.7 in the UK, we used WE8ISO8859P15.
> What should we use in 9iR2?
>

No need to change your character set. Presumably, you do most of your business in the UK and Europe... most of which would fit quite happily into a CHAR or VARCHAR2 using WE8ISOblahblahblah.

Should the day ever come where you want to sell something, occasionally, to Russia or Turkey, then you'd modify a few customer tables and change their CHARs and VARCHAR2s into NCHARs and NVARCHAR2s. And, assuming this is 9i, you would have had to have picked one or other of the unicode character sets for the national character set.. and you would probably have been wise to choose the ALUTF8 one (can't recall its proper designation off the top of my head). Because that way, all your non-Russian and non-Turkish customers, stored in the same NVARCHAR2 columns, would still only be taking up a single byte per character. If you'd picked the UTF16 one, you'd be storing the bulk of your non-Russian/non-Turkish customers in double-byte characters.

On the other hand, should your company decide to do MOST of its business in Tokyo in future, and occasionally still send a few trinkets back home to blighty, then you'd probably want to change the character set of the database itself to UTF16. The national character set is irrelevant at this point (it defaults to being the same as the character set, but it really doesn't matter). All your tables retain the use of CHARs and VARCHARs, and you don't bother using NCHARs/NVARCHARs. Being UTF16, all the Japanese custo mer details get stored as fixed-width double-byte (in UTF8, they'd be three bytes long). The occasional European customer also gets stored as double-byte characters, which is one more byte than they'd have used in UTF8, but since there aren't many of them, who's counting?

So much for theory: I reckon you're likely to fit the first bill. Stick with WE8ISO for your character set and make occasional use of a UTF unicode National character set when the need arises.

Regards
HJR
> Regards,
> Paul
>
>
>
Received on Wed Nov 20 2002 - 19:00:31 CST