In my setup, there's 5000 instances of Model, each one tagged with 5-9 tags
picked randomly from 10 available tags.
When get_by_model is executed with just two tags, the resulting query
passes too many Model PKs as a set for the IN clause of
TaggedItemManager.get_intersection_by_model.
Also, the result returned by get_intersection_by_model is
queryset.filter(pk__in=object_ids), which doesn't seem to be the most
optimal queryset that could be returned. I don't know if it's possible to
achieve anything more optimal with the current stable version of Django's
O/R Mapper.
Original issue reported on code.google.com by marcin.seredynski on 19 Feb 2009 at 9:25
Original issue reported on code.google.com by
marcin.seredynski
on 19 Feb 2009 at 9:25