Ramiro: I've set up some models / forms / formsets which demonstrate the change here:
http://dpaste.de/YhrI/ My initial report wasn't correct - the change in behavior is only seen when initalizing ModelFormSets . . . the ModelForm behavior is unchanged. With a formset, the value use to get initialized; now it doesn't. With a form, the value has never gotten initialized. I think 1.1 & 1.0.3 have the same behavior (no initialization in formsets); and SVN 10132 (1.1 beta) & 1.0.2 have the same behavior (initialization in formsets). On Aug 1, 6:56 pm, Ramiro Morales <cra...@gmail.com> wrote: > On Sat, Aug 1, 2009 at 8:28 PM, David Haas<david.h.h...@gmail.com> wrote: > > > Hi: > > > I'm hoping to get some feedback on a little change in behavior I'm > > seeing when initializing ModelForms between 1.1-beta & the official > > 1.1 release. > > > The quick and dirty explanation is this: I have a ModelForm which is a > > subclass of another ModelForm. The parent ModelForm has the fields > > attribute defined in its inner Meta class - the child ModelForm does > > not. In 1.1, when I pass in an instance of an object to initialize > > the form, only the fields appearing in the parent ModelForm's field > > attribute list end up with initial values - other fields are blank. > > In 1.1-beta, all fields (whether or not they're in the parent > > ModelForm's field list) get initial values. > > > In my case, I've worked around this new behavior by adding some code > > in the child ModelForm __init__ method to set the initial values of > > the relevant fields. I'm just curious if the change in behavior was a > > deliberate decision, and if not, does it merit a bug report? > > > And yes, I know that when you specify the fields attribute in an inner > > Meta class, only those fields get saved to the db. I have a system > > with separate data entry & review, and I'm seeing this problem on the > > review side. During data review, no matter what gets sent out, the > > only thing that I want to save in the database is "accept" or > > "reject", which lives on the parent form. > > I'm not seeing this, I've modified Django model_forms modeltest to > exercise what I interpret from your report and the new test passes > and shows identical behavior with both 1.1beta (r10133) and 1.1: > > http://dpaste.de/Yr0f/ > > Please modify these tests to describe what you are seeing if I've > missed anything. > > Regards, > > PS: Maybe you are not seein the difference in behavior at model > instance save time as per what's described in the docs in the note > right above this?: > > http://docs.djangoproject.com/en/dev/topics/forms/modelforms/#overrid... > > -- > Ramiro Moraleshttp://rmorales.net > > PyCon 2009 Argentina - Vie 4 y Sab 5 Setiembre > Buenos Aires, Argentinahttp://ar.pycon.org/2009/about/ --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Django users" group. To post to this group, send email to django-users@googlegroups.com To unsubscribe from this group, send email to django-users+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/django-users?hl=en -~----------~----~----~----~------~----~------~--~---
