Don’t get me wrong, I really enjoy working with Rails, but every now and then something about it bothers me. Josh Susser of has_many :through recently blogged about the sometimes differing functionality of size, length and count in ActiveRecord.

Did you catch that, _differing_. When using the count method on association collections ActiveRecord responds differently based on pervious usage. When you’re working alone or on small code bases this might not be a big deal, you might never run into it, maybe you’d even be able to use it to your advantage. The same could be said for larger projects or working with multiple developers, but it’s far more likely to cause a hard to track down bug.

I understand Rails is an open source project controlled by people with their own interests at mind, so please don’t read this as an attack against them or the project in general. I am not the first person to mention these kind of anomalies in the framework, and probably not the first person to have this thought, but would it make sense to fork the project?

I’m not sure it’s worth forking Rails to make it more “enterprisey” (read: consistent). It would be great to have a stable platform, maybe it would receive better acceptance in the office (ie: maybe my python coworkers wouldn’t pick on it). My biggest concern is that removing all these idiosyncrasies will make it impossible to port future creativity from the original authors. Again, that wouldn’t be the end of the world if such a project could attract new development, but I just don’t see that happening. Any thoughts?

I think I’m going to keep a little list going of the things that irk me about Rails. Each time I come across one Merb gets more and more tempting. I really want to dig into Merb, I’ve yet to get do anything more complex than a few basic tutorials, I just need a project…