Re: [SQLObject] multiple inheritance tree

2006-12-06 Thread Diez B. Roggisch
> > I'm digging into this - but if there is another solution, like creating > explicit table aliases to B and joining them with I got it working locally, and it passes all _my_ tests. The key was to prevent the childName-column to be overridden by the parents child-name column. So now, things wo

Re: [SQLObject] multiple inheritance tree

2006-12-06 Thread Diez B. Roggisch
On Tuesday 05 December 2006 23:05, Oleg Broytmann wrote: > (Returning to the list...) > > On Tue, Dec 05, 2006 at 10:42:41PM +0100, Diez B. Roggisch wrote: > > Lets consider this hierarchy: > > > > class A: > >pass > > > > class B(A) > >pass > > > > class C(B) > >pass > > > > class D(B)

Re: [SQLObject] multiple inheritance tree

2006-12-05 Thread Oleg Broytmann
(Returning to the list...) On Tue, Dec 05, 2006 at 10:42:41PM +0100, Diez B. Roggisch wrote: > Lets consider this hierarchy: > > class A: >pass > > class B(A) >pass > > class C(B) >pass > > class D(B) >pass > > class E(A): >pass > > > Now I have a page that shows all A,

Re: [SQLObject] multiple inheritance tree

2006-12-05 Thread Oleg Broytmann
On Tue, Dec 05, 2006 at 06:41:39PM +0100, Diez B. Roggisch wrote: > I have to paraphrase that: the real problem seems not to be that the support > isn't there - it it that the childName columns value isn't distinguishing > _all_ subclasses on the topmost class, as I was expecting. Consider this

Re: [SQLObject] multiple inheritance tree

2006-12-05 Thread Oleg Broytmann
On Tue, Dec 05, 2006 at 06:17:30PM +0100, Diez B. Roggisch wrote: > I just found out that multiple inheritance doesn't work over more than one > level. A program is beign written in our company uses 3-4 levels of inheritance, so I think you are wrong. Oleg. -- Oleg Broytmann

Re: [SQLObject] multiple inheritance tree

2006-12-05 Thread Diez B. Roggisch
On Tuesday 05 December 2006 18:17, Diez B. Roggisch wrote: > Hi, > > I just found out that multiple inheritance doesn't work over more than one > level. Is this an oversight, and people are interested in having a solution > for this, or is it by design? I have to paraphrase that: the real problem