6 hours on a train with Code Igniter

I’ve mentioned Code Igniter before but only now have I got round to really putting it through it’s paces by building something proper. I got the opportunity (if you could call it that) thanks to a six hour train journey from Oxford.

I’ve been thinking about sprucing up this site for a while, and have a few other personal projects that I really should get round to sorting out. Something I’ve noted of late is a move, amongst some parties, back to bespoke personal content management systems. Jeremy has been speaking about some of the feaures he has been adding recently, Christian had a few comments too plus Jonathan has been posting about his recent move to CakePHP from Movable Type.

I find this kind of interesting. Everyone started off with hand rolled solutions, seemed to move to the comfort of Textpattern or Wordpress and the more geekely inclined are on the move again. With such large user bases the big blogging platforms are (rightly) spending more time on bug fixes and user experience, rather than kick ass features. And us geeks want more Microformats and mashups that we can shake a stick at. Only this time we get to stand on the shoulders of giants by using frameworks such as Rails, Django or Code Igniter which means less time doing the boring stuff.

Short sales pitch. Code Igniter is cool, and more. The documentation is first rate and comes along with the download rather than just being available online (useful when your on the train.) It all makes sense as well which is nice, especially if you have some familiarity with the MVC pattern. It covers the basics so you dont have to – validation, database abstraction, session hanling. It’s also easily extensible through a nice plugin architecture and makes no assumpions about what you intend to build. I have a quick and dirty Textile plugin that I need to package up and put out – some discussion of this over on the forums which I may peruse.

In short expect to see some minor additions and the like round these parts as I try to transition away from Textpattern and any problems caused are no doubt the fault of some poor coding somewhere.

And I wonder what the next step will be? We have been through hand rolled to stable applications to frameworks to hand rolled based on frameworks. Any takers?