> > What I personally don't understand is: if all my databases
> > are UNICODE, why do I have to set the Client encoding to latin1
> > to get a correct result?
Karsten Hilbert wrote:
> Because LATIN1 isn't just a subset of UNICODE. If the data
> coming out of the database *is* UNICODE *and* your client
> *does* handle UNICODE directly you might get away with not
> setting a client_encoding, same goes for DB=latin1 +
> client=latin1. If, however, the DB delivers UNICODE but your
> client really wants LATIN1 you need to tell the database to
> convert the stored UNICODE to LATIN1 before delivery. That's
> what the client_encoding is for.
>
> Or so is my understanding of it.
>
> Karsten
Thanks, that's a pretty clear explication.
Only one question:
Wouldn't it be better if I just set my client encoding to UNICODE in stead of LATIN1? I suppose the UNICODE encoding
setis understood by windows (and delphi), since I write progs for a win enviroment.
Stijn.