> I think somehow, somewhere, the size of the default value is getting
> used instead of the size of the field itself. Weird. Is it specific
> to char(n), perhaps? That might help explain how the bug got past
> the regression tests.
I'm pretty sure it is specific to char(n), and it is due to the 4 byte
difference in length between the string and the storage. When we fix
this one we will also fix the "c char(2) default ''" problem too.
In fact, I could have sworn I already had looked at it. Oh well.
- Tom
--
Thomas Lockhart lockhart@alumni.caltech.edu
South Pasadena, California