Enter eZeAR

So what exactly is eZeAR? I have been searching for "The One True Web Application Framework(tm)" for about half a year now. I've looked at Symfony, I've looked at the Zend Framework... and yes, I have to admit, I even strayed from the pure PHP path and looked into Ruby On Rails. Of course, I didn't find TOTWAF(tm). Every framework had some things that I liked. And things I hated. So, like so many people before me, I just had to reinvent the wheel and roll my o... wait a minute, reinvent the wheel? I hate reinventing the wheel! There has to be a better solution! And there is. The Zend Framework, in its current Alpha state, really isn't a good framework. It can't do a whole lot of stuff. I totally hate its database layer. It has no commandline tools to ease application bootstrapping. Absolutely everything has to be done by hand, there's no grand design where you just drop files in, following a given naming scheme, and things magically just start to work. The coolest components in the framework, like the Lucene implementation, are even more like standalone components that can be used completely out of framework context. But it has potential. On the bright side, I really like the controller and view implementations. They're very, very basic. They just do the most needed things to be just barely usable. And they're not tied to any of the other framework components, like the ugly database layer or the input filter. Why is this good, you may ask? It is good because it makes the ZF the perfect metaframework - the "framework of frameworks" if you will. It is very easily extendable so you can drop in all the magic you want with just a few lines of code. It doesn't force you to follow its directory structure, file naming conventions or to use their favorite database access layer. It just allows you to do what you want and add all the magic you might need for yourself. Because of these unique attributes (and the nagging feeling that the ZF will face an enormous adoption rate once it's stable) I've chosen the Zend Framework as the basis of everything I'll write in the next few months. I'll open source all the components that I write myself to add the needed magic to make it a well-working framework for myself, and maybe others. In the next blog entries, I'll go into more detail and show off what I have already done and why I did it that way. Expect slow progress though - as I've said before, the time I've got for open source hacking still is rare. On the other hand, this also means having to find solutions that require the least possible amount of code (and thus, time) and maximum reuse of existing components - so it might get interesting