This may be a little tangential but I am wondering how many hobo users make use 
of DB constraints to supplement rails validations?

I'm not a SQL maven by any stretch, buy using an RDBMS without using 
constraints feels like paying for a DB without leveraging it's value.

Sent from my iPhone

On Feb 18, 2011, at 12:45 PM, "Donald R. Ziesig" <[email protected]> wrote:

> Hi Matt, et. al.:
> 
> Well I may have missed something in the addition of the invalid record and 
> the interaction with before_save, but I still can't get an error when the 
> child is invalid. (in the test app, there is no     invalid record saved, but 
> still no error message);
> 
> I created a test app with a parent table and a children table, both of which 
> initially had only name:string as columns (appropriate columns were added 
> later by hobo g migration).
> 
> added to parent model:
> 
>   has_many :children
>   children :children
> 
>   validates_presence_of :name
> 
> =====
> 
> added to child model:
> 
>   belongs_to :parent  
>   validates_presence_of :name
>   
> =====
> 
> added to children controller:
> 
>   auto_actions_for :parent, [:create]  
>   
> ============================================
> 
> Not much to type ;-) 
> 
> performed hobo g migration
> 
> =====================================
> 
> Performed the following tests:
> 
> =====
> 
> Added parent with no name:
> 
> Got big red box saying Name can't be blank.  No parent added.
> 
> =====
> 
> Added parent "Mama"
> 
> got to the show page for Mama, with Add Children box:
> 
> Pressed "Add" with a blank Name field.
> 
> Got an empty "Child" page with NO ERROR, but with a link back to the parent 
> page.
> 
> Followed the link to the parent page.  Mama had NO children - so even though 
> I got a blank child to look at (in the other apps, which had more columns, 
> they were displayed with the values I entered), the record was not saved.
> 
> =====
> 
> Added child "Suzie",  pressed "Add"
> 
> Got Suzie's child page.  went back to Mama and Suzie was there.
> 
> =====
> 
> I don't know if attachments will get through to the list, but the tgz file is 
> less than 200K and is ready to run
> 
> =====
> 
> I'm going back to look at the before_save issue in my primary app     (its 
> complicated enough that I probably messed it up somewhere or     
> misinterpreted the results).
> 
> Thanks,
> 
> Don Z
> 
> On 2/18/2011 11:09 AM, Matt Jones wrote:
>> 
>> On Feb 18, 2011, at 11:02 AM, Donald R. Ziesig wrote:
>> 
>>> I hacked around the addition of the invalid record - I don't remember 
>>> seeing this in any documentation so I worry about the future.
>>> 
>>> It appears that returning false or nil from before_save when there is an 
>>> error stops the record from being saved YMMV.
>>> 
>>> I still can't make the error message(s) appear though.
>> This definitely shouldn't be happening - I've got apps in production that 
>> use validations along with auto_actions_for and they work fine.
>> 
>> The fact that you're *getting* to the before_save callbacks indicates that 
>> something else is disabling validations, as a failed validation should 
>> cancel the save before it even gets to those...
>> 
>> --Matt Jones
>> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Hobo Users" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to 
> [email protected].
> For more options, visit this group at 
> http://groups.google.com/group/hobousers?hl=en.
> <parent_child_test.tgz>

-- 
You received this message because you are subscribed to the Google Groups "Hobo 
Users" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/hobousers?hl=en.

Reply via email to