On Thu, 2009-02-26 at 11:08 -0500, Calvin Spealman wrote:
> I'm looking for the best way to query based on the existence of an
> instance of a subclass. In my case, there are Actions that have
> several subclasses, like FriendRequest, PhotosTagged, etc. The user
> has selected a set of types they
On Thu, 2009-02-26 at 12:56 +0100, Killian wrote:
> Anyway, I'm wondering what's the use of explicitly indexing the
> objects in an unordered queryset? When it's unordered, the result
> you'll get is random anyway, you cannot predict the outcome at all.
> The only reason I see is to select multipl
I'm looking for the best way to query based on the existence of an
instance of a subclass. In my case, there are Actions that have
several subclasses, like FriendRequest, PhotosTagged, etc. The user
has selected a set of types they are interested in, so I need to query
for all the actions that are
Anyway, I'm wondering what's the use of explicitly indexing the objects in
an unordered queryset? When it's unordered, the result you'll get is random
anyway, you cannot predict the outcome at all. The only reason I see is to
select multiple 'random' objects, which is much easier solved by
order_by
Hello,
Bigint patch probably should be obsoleted by current state of easy new
field definitions. I have set patch state from 'accepted' to 'design
decision needed' before 3 months and there is still no response from any
of core developers. I will be very glad to see any decision in this case.
On Feb 25, 10:08 pm, Jacob Kaplan-Moss
wrote:
> See, the reason I closed your ticket as wontfix is because I don't
> think there is. That is, I think it's impossible to make
> ``unordered_qs[N] == unordered_qs[N]`` 100% of the time without (a)
> adding an artificial ``ORDER BY`` clause or (b) mat
