Russell Keith-Magee wrote: > On 1/27/06, Robert Wittams <[EMAIL PROTECTED]> wrote: > >>Russell Keith-Magee wrote: >> >>>2) I have a minor problem with the _set suffix: to me, _set implies >>>uniqueness in the returned results, which will not exist unless >>>.distinct()/distinct=True is used. >> >>Could you provide an example where you would actually end up with >>duplicate results here? I'm having trouble thinking of how this would >>actually occur. AFAICT, each child object will only show up once in the >>results. If there is one, I agree that it would make sense to use >>distinct by default. > > > Any query involving joins over a m2o or m2m relation; if there are > multiple matches in the related table, there will be multiple rows > with the fields from the primary table: e.g., > > daily_planet = Newspaper.objects.get_object(name='Daily Planet') > daily_planet.reporter_set.filter(article_headline__startswith="Foo") > > will duplicate results for every reporter that has written more than > one article with a headline starting with 'Foo'.
I agree we don't want this current behaviour. I had some hazy idea that this got turned into group by malarkeys - obviously not... Yes, I think distinct is the right way to go here.
