Nostalgia warning: I also grew frustrated with Rails / ActiveRecord and jumped to Merb / DataMapper with excitement. Perhaps it will sound like a small thing, but I remember feeling especially frustrated that every ActiveRecord model became such via inheritance. This drove me crazy:
class User < ActiveRecord::Base
Why should my `User` extend from `ActiveRecord::Base`? Is a User of my app also an ActiveRecord::Base? What is an ActiveRecord::Base anyway?
So, when DataMapper came along and favored composition over inheritance I was sold.
However, my experience with DataMapper was that it didn't have the stability of ActiveRecord, nor did it reach feature parity.
And, then Rails killed Merb. At the time I thought it was a tragedy. The competition Merb provided did improve Rails, it would have been nice if that competition had been longer term, though. On the flipside, I wonder if there would have been enough community to support two large Ruby web frameworks.
In the end I made peace with Rails and ActiveRecord (and, believe it or not, even ActiveSupport!). I don't have great love for Rails, but it is a powerful tool, so I accept it for what it is.
That said, I think the author's criticisms are well stated and hopefully there will be some influence on the future direction of Rails, etc, and the way devs approach Rails.
So, when DataMapper came along and favored composition over inheritance I was sold.
However, my experience with DataMapper was that it didn't have the stability of ActiveRecord, nor did it reach feature parity.
And, then Rails killed Merb. At the time I thought it was a tragedy. The competition Merb provided did improve Rails, it would have been nice if that competition had been longer term, though. On the flipside, I wonder if there would have been enough community to support two large Ruby web frameworks.
In the end I made peace with Rails and ActiveRecord (and, believe it or not, even ActiveSupport!). I don't have great love for Rails, but it is a powerful tool, so I accept it for what it is.
That said, I think the author's criticisms are well stated and hopefully there will be some influence on the future direction of Rails, etc, and the way devs approach Rails.