>>Jon, what is the reason for not specifying all columns in the INSERT? That is the standard way and then you don't have to worry about mystery values getting in.
>
>Mike,
>
>I understand where you're coming from on this, and it's certainly a valid way to resolve what we've been seeing.
>
>The main argument I see against this is that if the table in question ever has a structure change, then we would have to make at least one source code change to take into account the structure change, which isn't a very dynamic way to handle things. Given how I've always understood the INSERT command to work, you should not need to specify all fields in the field/value list, because the default values should kick in for all fields that are not specified. If I would be specifying all fields all the time, then I feel there's really no point in supporting default field values.
>

If it has a structure change aren't you going to want to populate the new field(s) on an insert?