--- In firebird-support@yahoogroups.com, "peter_jacobi.rm"
<peter_jacobi@g...> wrote:

> Hi Mark,
>
> --- In firebird-support@yahoogroups.com, "mbellisle_retire"
> > I must have just woke up. The problem is that I'm using the

WIN125x

> > character set.
>
> Are you really seing clear now?
>
> There are three character sets to consider:
>
> 1. The database column's declated character set
> This seems to be NONE in your case, which is always
> a problematic choice, when non ASCII data is to stored.
> Unfortunately it's the default.
>
> 2. The actually used character set in a database columns
> If you have a byte value 0x99 in your data and it should
> read <TM-symbol>, it's a hint that this CP1252 data
>
> 3. Your connection character set, set in ISQL with
> set names
> When 1. is NONE 3. should also be NONE and you are
> expected to handle character set issues yourself.
> If 1. is not NONE, 3. should not be NONE. Firebird will
> do the transliteration for you. Typically the bot character
> sets are choosen to be equal or one of them is UNICODE.
>
> 4. The character set your client uses to display and
> interpret strings. If this doesn't equal 3., all
> sorts of funny chars get displayed. For ISQL this
> character set is set by chcp.
>
> Peter

Well I suppose my vision is still foggy. I haven't worked with IB a
great deal, so I'm not familiar with all the nuances of dealing with
selecting the character set. If I understand you correctly, your
saying even though the db doesn't spec a char set, I should choose
UNICODE, just to ensure some level of correct translation of the
characters, if I come across any others?

Thanks very much for your help.
Mark