> as a followup, here's some more confidence inspiring behaviour from
MySQL
> (my version is what shipped with RH 7.2, 3.23.41.
>
> mysql> create table testg (t1 numeric(30,4));
> Query OK, 0 rows affected (0.00 sec)
>
> mysql> insert into testg values (12345678901234567890.1234);
> Query OK, 1 row affected (0.00 sec)
>
> mysql> select * from testg;
> +---------------------------+
> | t1 |
> +---------------------------+
> | 12345678901234567168.0000 |
> +---------------------------+
Foxpro has a very similar philosophy in design. It will do many awful
things to your data, like silently truncate your strings, etc. It's a
pretty safe bet that the developers at Bruce's company are quite used to
dealing with these issues in application logic code. They are probably
also used to fixing busted indexes and repairing data and all the other
things that come along with the direct path to the data. Mysql is
actually a lot like foxpro in a lot of respects, although it lacks a lot
of the programming power of foxpro. Moving to this system will not gain
a whole lot by any measure except to allow easy connectivity over the
internet.
Postgres, OTOH, offers a world of advantages, both technical as well as
the PHB type stuff.
Merlin