Re: [Renaissance]platform objects' awakening

[David Ayers]

Dirk Lattermann wrote:

> Hello Nicola!
>
> Thanks for your instant help!  I'm making some promising
> advances with the EOAssociation connector.  Binding a text field
> to a database via an EOControlAssociation is working here now!
Wow!  Great news!

> I haven't factored out the new classes into a seperate library,
> but will do it after I have implemented some more.
>
> When trying to make EODisplayGroups available to Renaissance, too,
> I stumbled over the following:
>
> EODisplayGroups have an property fetchesOnLoad which, when set to YES,
> tells them to issue a fetch to their EODataSources after they have
> loaded from a nib (resulting in a fetch from the database).  In Mulle's
> implementation of EOInterface, this is done in
> -[EODisplayGroup awakeFromNib], I don't know where it's done in
> Apple's version.  To be able to use this feature, I suggest the
> following patch to GSMarkupAwaker.m.  That way, the platform object
> can be a real EODisplayGroup and not a subclass which implements
> -awakeFromGSMarkup.  Yet I don't know if that would harm some other
> platform objects.
How about adding a category to EODisplayGroup in the library to be that 
implements
@implementation EODisplayGroup (Renaissance)
- (void)awakeFromGSMarkup
{
 [self awakeFromNib];
}
@end

That way, if we ever need to fo something special for GSMarkup, we can 
just alter this implementation accordingly.  I don't know what the side 
effects of allways falling back to awakeFromNib within GSMarkupAwaker 
may be, but it feels dangerous.

Thanks for your effort!  I hope to have that EOEditingContext patch 
ready soon, to fix that bug, but it's quite intricate and I need to fill 
a lot of holes.

Cheers,
David