Thanks for the reply, I used a trial of Database Workbench 5 to create the
new database rather than the paid for version 3.4.4 I usually use (just
testing to see if I want to upgrade) and I checked all my existing databases
and this new database was created to use UTF8 (by default I guess and I
didn't change it) while all my other databases seem to use a characterset of
NONE. I'm still not entirely sure what exactly the problem was because I
tried deleting a few records and I still couldn't add a record or edit the
artist and title fields of any existing records. So I created a new database
with Database Workbench 3.4.4 and created the tables, generators and
procedures using the DDL statements from the old database (editing out
references to UTF8) and then used the datapump to copy data from the
existing database to the new one. After renaming both databases everything
appears to be working properly again.

I looked up UTF8 and the Wikipedia article says "UTF8 is a character
encoding capable of encoding all possible characters (called code points) in
Unicode" so if that's the case then I still don't understand exactly what
the problem was with the old database; why I was able to enter 104 records
and then suddenly start getting a -303 error when adding a record or editing
the Artist or Title fields of an existing record? Also what's the real
difference between having a database with a default character set of UTF8
and NONE? I'm asking for future reference and because I'm pretty sure that I
didn't fix the problem but instead just developed a workaround and I'd
really like to know?

>Michel Le Clezio
>
>Take a look to charset used in text and the Charset of database... May be

in one record you use is an unsupported char... by the Charset of
database... this can happen if you paste

> a value... ( a value not build by your keyboard)
>
>
>Le Dimanche 15 mars 2015 5h28, "'Chris LeFebvre' lefebvrec@...

[firebird-support]" <firebird-support@yahoogroups.com> a écrit :