[bithub]

At a quick glimpse at the framework code, I found several no gos for me: 1) First file i opened contained Methods with > 600 lines of code with mixed variable naming (camelcase and underscore). I don't want to debug that stuff, if I run into a framework bug. 2) Several classes/interfaces in a single file. What if I ie want to write my own adapters by using the given interfaces?

On the other side the code is very well commented (not all but most of it), which is a very huge plus.

[EGreg]

Thanks for the feedback, but are you looking at the right framework? I am hard pressed to imagine how you could encounter either 1) or 2) in it, let alone both in a single file. As for your question, PIE is designed to let you write any adapters you want. For example it currently supports only MySQL, but you are free to add support for e.g. Postgres, SQLite or even NoSQL databases.

I'm just a bit surprised by your comments because they describe the exact opposite of how most of PIE is designed, and the philosophy behind it. Although I take your point with mixed variable naming. There is lots of room for standardizing the internal codebase more than it is.

[jameskilton]

framework/classes/pie.php 1082 lines long

Methods existing in the file: microtime, exceptionHandler, includeFile, autoload, var_dump, do_dump.

It's the first file I opened, and I see exactly what this guy saw immediately as well.

[EGreg]

Right ... but I thought he said there were methods with > 600 lines of code, and files with multiple classes/interfaces all over the place. That's why I took issue with it. But yeah, the Pie class has all the core functions.