Ciao,
On Tue, Mar 17, 2015 at 4:51 PM, Little, Doug C
<Doug.Little@vend.frb.org> wrote:
> However, I could use a generic wrapper - but I don't think the customer would like this.
That's a solution, but I still have some doubts about the final aim.
First of all, for a successful statement there is sqlstate = 0 and no
exception. And what is the point of seeing a successful statement?
Does it suffice to know if the transaction has been committed or not?
That would be a lot easier, for example, from a program point of view.
Similarly, knowing the sqlstate of a wrong statement what information
adds for later analysis (because storing it into a table means that to
me)?
I believe there is something unclear in the design of the whole idea.
However, I would go for a wrapper.
Luca
