>> I wrote:
>> Firebird's Unicode-FSS
>> (aka UTC-8) checks for a byte-length overflow, but ignores character
>>length overflows. One argument is "If the string fits in the space, why
>>throw it out?" In the case of our system tables (nominally Unicode-FSS)
>>that's really important or every name would be limited to 10 characters.

Arno Brinkman wrote:

> I think we need to solve this, but reading your post it's not that easy?

My guess is that we could put in a hack that changes the handling of
user data in Unicode-FSS, but that the reaction will be strong and
negative from people who've used it since InterBase 4 or 5 and count on
the fact that their western European data actually averages about .8
characters per byte. I know the next answer. We'll have a
configuration parameter to allow or disallow overstuffed Unicode-FSS
fields...

Sounds of Ann running screaming out of the room, leaving hanks of
torn-out hair in her trail...

Cheers,


Ann