I've been brainstorming (ie, drinking more coffee than I should), and
what I came up with to be the best (IMHO) solution for A) Less
confusion, and B) Less risk of API breakage is:
ModelForm.is_valid(include_all_fields=True)
or
ModelForm.is_all_fields_valid()
and neither are going to be an issu
On Sat, Apr 3, 2010 at 3:03 PM, orokusaki wrote:
> Russ,
>
> I think you're 100% right, and the "wrong place" part hit the nail on
> the head. This morning I got really frustrated because I couldn't
> quite see the big picure yet pertaining to the ORM and it's
> relationship with ModelForm, partly
On Sat, Apr 3, 2010 at 8:14 AM, Boris Schaeling wrote:
> On Thu, 01 Apr 2010 11:11:47 +0300, Russell Keith-Magee
> wrote:
>
>> [...]I don't deny that it would be *really* nice to be able to
>> automatically call full model validation on a model on form save - the
>> problem is that we can't do th
Russ,
I think you're 100% right, and the "wrong place" part hit the nail on
the head. This morning I got really frustrated because I couldn't
quite see the big picure yet pertaining to the ORM and it's
relationship with ModelForm, partly because there is so much going on
with state changes and you
On Thu, 01 Apr 2010 11:11:47 +0300, Russell Keith-Magee
wrote:
[...]I don't deny that it would be *really* nice to be able to
automatically call full model validation on a model on form save - the
problem is that we can't do that while retaining backwards
compatibility.
How about a setting
On Sat, Apr 3, 2010 at 2:51 AM, orokusaki wrote:
> On Apr 2, 2:00 am, Russell Keith-Magee wrote:
>
>> The broad goal is certainly reasonable and desirable. It's really a
>> matter of finding a way to make it work that doesn't involve
>> completely breaking (or disfiguring) the API that we already
On Apr 2, 2:00 am, Russell Keith-Magee wrote:
> The broad goal is certainly reasonable and desirable. It's really a
> matter of finding a way to make it work that doesn't involve
> completely breaking (or disfiguring) the API that we already have.
>
> Yours,
> Russ Magee %-)
I've been working on
On Fri, Apr 2, 2010 at 6:25 AM, orokusaki wrote:
> Hey Russ,
>
> I'm not on the model.full_clean stuff anymore, and I apologize for
> burning so many cycles on that point when you're in the middle of 1.2
> dev. I'm just wondering if what I proposed above sounds reasonable.
> The form validation tu
Thanks Jacob, I'll give that a try.
On Apr 1, 7:44 am, Jacob Kaplan-Moss wrote:
> On Thu, Apr 1, 2010 at 3:11 AM, Russell Keith-Magee
>
> wrote:
> > Melodrama aside, as we've told you before, the docs clearly say that
> > full_clean() isn't called by form.save(). The docs also give you the
> > r
Hey Russ,
I'm not on the model.full_clean stuff anymore, and I apologize for
burning so many cycles on that point when you're in the middle of 1.2
dev. I'm just wondering if what I proposed above sounds reasonable.
The form validation turned off completely for fields that aren't
included makes it
On Thu, Apr 1, 2010 at 3:11 AM, Russell Keith-Magee
wrote:
> Melodrama aside, as we've told you before, the docs clearly say that
> full_clean() isn't called by form.save(). The docs also give you the
> reason - backwards compatibility.
>
> I don't deny that it would be *really* nice to be able to
On Thu, Apr 1, 2010 at 9:11 AM, orokusaki wrote:
> I'm working on an SAAS project, and there is an ``account`` attribute
> (foreign key) on every model in the project (similar to those who have
> a ``user`` or ``created_by`` attribute on every model). ``account`` is
> added to the request object u
On Thu, Apr 1, 2010 at 2:17 PM, subs...@gmail.com wrote:
> Seems like a security hole, whereby people may supply additional
> fields if they can guess their counterparts on the model. Its
> 'exclude', not 'exclude_maybe'.
Please be *very* careful about using the words "security hole" - those
word
Seems like a security hole, whereby people may supply additional
fields if they can guess their counterparts on the model. Its
'exclude', not 'exclude_maybe'.
...Unless I'm missing something fundamental.
-S
On Mar 31, 9:11 pm, orokusaki wrote:
> I'm working on an SAAS project, and there is an `
Let me just say that my non-patch above is just an abstract idea, and
I don't know if it will work like that without other changes, but I
think it gets the idea across.
--
You received this message because you are subscribed to the Google Groups
"Django developers" group.
To post to this group,
I'm working on an SAAS project, and there is an ``account`` attribute
(foreign key) on every model in the project (similar to those who have
a ``user`` or ``created_by`` attribute on every model). ``account`` is
added to the request object using a MiddleWare class.
When I'm writing views, I have t
16 matches
Mail list logo
