Luis and his team of elves have been busily preparing Coldbox 2.5 and it looks like they managed to ship a little early for Christmas this year.

Check it out!

I just wanted to point to some of the great work that Kevin Roche has done on the Fusebox scaffolding. I know that there are lots of other frameworks out there, but Fusebox is definitely worth checking out. It doesn't require OO developers, provides good MVC separation, and the latest version it has some very cool magic from the curmudgeon himself that makes it worth checking out the underlying code even if you don't use the framework for some of your projects. Who knows, it might even save your job!

In case you don't know, Rails 2.0 is now out. I'm actually not a huge Rails fan. It's opinionated software and I'm an opinionated programmer - things just wouldn't work out between us!

That said, skimming through the feature lists I've already found some things I did ages ago, some that I was planning to do now, and some that I'll be adding to my task list as I think they are great ideas.

It won't tell you which one to use for your use case, but at least wikipedia has a decent list of MVC web application frameworks - including most of the main ColdFusion ones. Check it out.

One of the many stories I didn't get a chance to mention recently was the great efforts by Aaron Roberson and Luis Majano. LightWire is now a supported DI engine in Coldbox.

Great work guys - many thanks!

Brian Rinaldi just did a great post dissecting how Mach-II actually works. It is only part 1 so it focuses on the load up process, but it is a great resource and puts all the nice new 1.5 features in context.

Only problem is that it was a lot of work, so if you want him to complete the series, please comment and let him know!

I've been rebuilding my framework (again). The last cut was pretty good, allowing a small number of simple methods to specify most of the functionality for common web applications declaratively. It used a lot of automagic and convention over configuration to allow a very DRY set of metadata to create applications quickly and easily. But as I played with it I started to feel something wasn't quite right . . .

[More]

Sam picked up on an interesting article and Sean put together a great response. I just wanted to look at the discussion from the perspective of Domain Specific Languages to provide another way of looking at such issues . . .

[More]

This article is a really nice comparitive review of a number of different Python frameworks (both full stack and more focused). Even if you have no interest in Python, it provides a really good overview of some different approaches to thinking about web application frameworks.

So, who's going to write the CF companion?!

While I don't agree with every point, I think the Web Framework Manifesto gives a good sense of some of thing I expect to see in a framework (and plan to include in LightBase). Some of the "independent UI components" while still allowing the back button to "do the right thing" is a pain to implement in a language without closures and a server memory hog in any application, but they are all things I'm pondering for version 2.0 of LightBase (i.e. I'm considering them but not coding them yet). Right now I'm focused on the basics like a DI engine and ORM rolled just the way I like it as well as my page controller model, composable views, automated handling of mapping of form fields for single or multiple object forms and generation of "first cut" of everything so you can just extend and customize if you want.

I think it is safe to say there is still plenty of room for both existing frameworks to extend the scope of what they do and for new frameworks exploring different metaphors. I think there is also a place for frameworks that keep true to a central vision and (for instance) stay solidly in the MVC space. DHH noted that Rails is not the "one true framework" as the more problems your framework solves, the less suited it is to each problem, so frameworks optimized for subsets of use cases will make more sense for people whose use cases consistenly fall within those spaces whereas more generalized frameworks will fit better for people with a wider range of needs who are prepared to spend a little more time on each project in return for a framework that provides them with more flexibility.

I think the next couple of years are going to be really exciting times as we look at all of the frameworks (especially outside of the CF world) and move the CF frameworks world to the next level.

What are you working on or hoping for in terms of ColdFusion frameworks and features?