gpuds framework: reflections on my software developer life cycle

On my last trip out to LA for work, I spent some time on the plane brainstorming on the data model and abstraction for gpuds.  I basically want to take the core feature set in the current incarnation of gpuds and abstract it into a "real" ColdFusion application.

I'm starting to realize that most of the ColdFusion development I've done in my career, starting back with CF 4.5 (I think), hasn't really evolved with the evolution of ColdFusion itself.  I'm still writing html/page based apps that require a lot of tedious maintenance and aren't very scalable.  In working more and more with BlogCFC and Mango Blog, I'm realizing some of benefits of CF 7 and 8. 

I think that I knew all along, in theory, how great CFMX was, but never from a practical perspective until I started working on the gpuds framework.  I can't even remember when CFCs were introduced, but aside from facilitating communication from the Flash player to the server, I think I've really been ignorant on why they are so important.

I don't think my applications were ever bad, but they could have been better.  The nature of ColdFusion as a Rapid Application Development platform is what perpetuated my continual use of less-than-efficient application design.  My applications worked and I could crank them out quickly, so why would I bother with any other approach or even framework?

I also think the fact that I'm doing much less development now and more project management is why I'm able to see these things.  I always had a deadline or projects piling up, leaving little time for analysis or post-mortem.  Now that I see these things from a different perspective, I can see all the shortcomings of my approach.

It's amazing to me that I've "discovered" all this in just the scope of working on my gpuds framework, especially considering that I've only just begun!