The late half of 2008 was particularly busy time for me, and between all the conferences and awards programs, not to mention all the stuff happening in my own business, I didn’t end up having a lot of time for actual coding.
The last few days I’ve had a quiet sit down for the first time with Fusebox 5.5. It’s been an interesting experience and while there’s some things I like, there’s also some things I’m not so fond of. I’m going the No XML route – not because I thought the XML was in some way bad, but because writing ColdFusion code seems more natural when its, you know, in ColdFusion.
Helping me in my explorations were Jeff Peter’s book How to Drive Fusebox 5.5 and Sean Corfield’s Cat Club sample applications. Jeff’s book was handy for explaining all the six billion possible configurations for Fusebox apps these days, and where I had questions that the book didn’t answer, Sean’s sample apps jumped in and showed me things like how the different levels interact with the new CFC circuits and fuseactions.
It’s probably too early for me to be making an sweeping generalisations, but in short: what I like is the implicit circuits, use of Application.cfc, and assigning the results of model operations to variables within the controller (at least when using CFCs for circuits). I was never comfortable with the model layer simply creating variables available in the current request without passing them explicitly back to the controller. This may be a stylistic thing that was possible within FB4, in which case please discount this little gush.
What I don’t particularly like is that there are a number of things that seem rather repetitive and tedious. But I’m sure I’ll work out what’s required and what’s not soon enough.
I read today on Adam Haskell’s blog that he’d taken some time off but is now back working on Fusebox. Which is great news – it’s good to see forward progress, and I can’t wait to see how it will be improved even further.