On Jan 30, 2010, at 1:05 PM, Kent wrote:
> While its certainly possible I've done something that isn't designed
> for, I am getting inequivalent SQL statements when I flip to
> use_ansi=False (unfortunately, I need to support oracle 8i).
I've committed a fix in r6712 which changes how the traver
On Jan 30, 2010, at 1:05 PM, Kent wrote:
>
> p__a, od__a, p__b, and od__b are aliases I've defined in order to work
> out this fairly complex query
>
> These joined aliases are being added to several of the outer selects
> where I didn't intend for them. This is causing a cartesian product
> a
While its certainly possible I've done something that isn't designed
for, I am getting inequivalent SQL statements when I flip to
use_ansi=False (unfortunately, I need to support oracle 8i).
I may try duplicating the problem with a more simple query so I can
post the tables & mappers here.
This i
