With vBulletin 4.0 blindly around the corner, where does that leave developers? If we start something big, by the time it is ready, it’ll be about time to rewrite the whole thing. Or, if we’re lucky, it will work for a short while, but we’ll have to recode it all shortly after.

The major problem here is the dependency on vBulletin. Since we’re making things for vBulletin, it doesn’t seem like much of a dependency, but it is. One minor change can render a program useless. Having the fate of our work rely in someone else’s hands isn’t the most reassuring thought. Then there’s the whole template editing nightmare.

The other night I was thinking about this, and the idea was to use other frameworks, and only integrate the user authenticate system.

It’s actually a pretty easy task to do - it took me about 5 minutes on cakePHP, granted I already know cake quite well. I also managed to do it for CodeIgniter for a friend of mine, and that took a bit longer — 2 hours. The key is to use the same database (and prefix) as vBulletin, and just use vB data as you need it. You can even use vBulletin’s cookies.

After having used other frameworks, and even writing several of my own smaller CMS systems from scratch, I can conclude that developing for vBulletin isn’t all I used to make it out to be. While I’m sure we will see MVC for vBulletin 4.0, the current state of things for large projects is an absolute mess. This is evident when looking at Jelsoft’s own modifications. The product system is not powerful enough. So, naturally, we are turning to other solutions.

I am going to start using Cake or CI for my future projects.