Lots and SQL backend

[Phil Longstaff]

Derek Atkins wrote:
> Phil Longstaff <plongstaff at rogers.com> writes:
>
> [snip]
>   
>> Once lot->splits == NULL, there are no splits, so unless there is some
>> other way that the payment might be tied in, it would be safe to
>> replace "lot->account = NULL" with "qof_instance_set_destroying(lot,
>> TRUE)".
>>     
>
> No, I do not think it would be safe to destroy the lot from under the
> invoice, at lease without auditing all uses of the API.
>
>   
ok
>> Phil
>>     
>
> -derek
>
> PS: I'll just note that this is why I keep saying that the DB backend
> is just a data store; putting in table requirements like this just
> causes problems for GnuCash and just makes everyone's life more
> difficult.
>   
Well, the other option was to remove the table constraint.  I had hoped 
to leave it in, but the code would need to be scrubbed and I don't have 
the time right now to do it.

Phil