On 30.08.2016 o 08:14, Stefan Heymann lists@...
[firebird-support] wrote:

>> One of the legacy databases I happen to maintain has character sets and
>> collations messed up. [...]
>
> For a database that is messed up like this, I'd go the clean way and
> use a pump. That would also give you the opportunity to change it to
> Unicode (UTF-8) at the same time, which is a transition that you will
> have to do anyway.

Probably you're right and that's how it's gonna end. Nevertheless, it
would be nice to know the inner workings of character sets and
collations (and the corresponding flags in system tables), what kind of
manipulation is possible on existing data and what impact it might
cause. That kind of knowledge might be valuable for its own sake, so if
anyone is aware of any sources of information casting some light on the
subject, please point them out. Knowledge never hurts ;)

thanks and regards
Tomasz