Eloy Duran

not sure I like the idea of silently eating the f.fields_for block when the association is nil.

Like discussed in #rails-contrib, this is not silently eating ...

Very nice, thanks a lot!

Returning an empty string from fields_for_with_nested_attributes feels dirty for some reason I can't qualify, but it does communicate ...

Do we need an additional check on the version of libxml that's installed? This is with a stock ruby install from 10.5.6

Yes I think we do. The version that sh...

@ Di Marcello: I quickly skimmed it and I understand the intent. But there are no tests and I want extensive test coverage.

Maybe, now that I understand how you...

@Lawrence: If you really need to change the behaviour of other methods you can override them in a Module and then call super. This is done whit the has_many, ha...

Invoking destroy will run ActiveRecord::Base callbacks, cause record is an instance of that class and doesn't know to belonging to an association.

Sorry, I do...

Yes, you got it right. And it should definitely be mentioned in the documentation.

About the extra work necessary in the controller; you might want to shove tha...

I agree we should fix this error people run into. But I wonder if simply using a regular fields_for is what the user expects. For instance, how about dynamic de...

Oh and I forgot to add that it is weird that validates_associated would fix this…

Whats the best (fastest) way for me to do that?

The fastest way to find where a bug was introduced, if it was, would be to use git-bisect. You create a file w...