Re: upgrading from 5.0 to 8.0.5

Hi Tuomas,

For most European countries the characterset should be WE8ISO8859P1. This is the Latin-1 alphabet. There are two ways to change a characterset. The official one: create a new database. The unofficial one: update the table prop$ connected as SYS.

Hth,

Sybrand Bakker, Oracle DBA

Tuomas Hosia wrote:

> Hello everybody!
>
> I have Oracle 5.0 under DOS and I'm trying to upgrade it to 8.0.5
> under Linux.
>
> I made an export file from 5.0 and I can import it into 8.0, that's
> not the problem.
>
> The problem is character set, especially scandinavic special
> characters ä and ö.
>
> IMP's argument CHARSET does have effect,
> but best result were that umlaut was simply stripped, with charset
> setting WE8PC850.
>
> On the other hand, if I insert a row (into 8.0) which has these
> characters, they appear as '?' when selected. ( Same effect both on
> console or via sqlnet &sqlplus 8.0 on NT. )
>
> select dump(field) clearly indicates that characters are stored as
> '?' and naturally there's nothing you can do about it then.
>
> I have 8.0.5 (pre-release) database up with default starter database
> and I think I made a mistake in there and the database won't accept
> national characters ( means that is in US7ASCII, I think? ).
>
> I suppose there's no way to change the character set of a (8.0)
> database?
>
> I could always drop old database and create new with proper character
> set but includes a lot of work, so any ideas are welcome.
>
> Tuomas
> --
> hosia_at_lut.fi(Tuomas Hosia) DoD#1684 \ On paljon hauskempi olla pakanana,
> Lappeenranta University of Technology \ ei tarvitse koskaan olla vakavana.
> Kotisivu: http://www.lut.fi/~hosia/ \ -o Eppu Normaali o-

Received on Sun Jan 17 1999 - 15:18:12 CST