The Future of Waffle

I have big plans for PHP Waffle, and this is how (I hope) things will go. Naturally, the Roadmap is subject to change. This is at least what I would like to accomplish for the project:

0.4 - Due ???; Status: In Planning
0.3 - Due 2nd Quarter 2004; Status: Completed (June 2005)
  • Implement the menu system. (Canceled, see the developer log for details)
  • User authentication (in CVS) and permissions (postponed until 0.4)
  • Application Switching (in CVS)
  • Unix/Linux/Shell installer script (implemented in version 0.2)
  • Configuration Control Panel (postponed until 0.4)
  • WYSIWYG Editor (in CVS)
  • XML Form Configurations (basic)
0.2 - Due February 2004; Status: Completed (May 2004)
  • A fully-functional application layer for the framework.
  • Construction of the Module Manager application.
  • Implementation of the Control Panel sub-system.
  • Improved "document" handling.
0.1 - Due January 2004; Status: Completed
  • A proof-of-concept web-based file manager.

Additional plans will be made in the future.

Here is my TODO list, features to add to Waffle:

Organized Cache
After having installed Waffle on my first Windows server, it became clear that the generated data/cache files should all be in one directory. This is because Windows has no way to delete a non-contiguous list of files/directories easily.
Auto-determine Dataobjects
Ruby on Rails proved XML configuration files are pointless and annoying — especially when it comes to defining database tables. Waffle suffers from this, and it needs to be fixed.
Less Config, Less XML
Again, what Rails has shown is that less is more. Good design makes configuration unneeded. I know this now, and Waffle will change for the better because of it.
Purge QuickForm
This is already started. PEAR QuickForm is/was Waffle’s weakest link. It sucks. I cannot be more annoyed with anything in PEAR as I am with this module. The replacement already works, but I will need to perform an exorcism on the system to remove any trace of this scum.
Full UTF Support
My hosting provider is forcing me to stick with MySQL 4.0.x, but that should not prevent me from getting a real UTF version of Waffle ready.
Improved UI
The current UI is OK. It works well for moderate to advanced computer users. My mother would have trouble with it, and so a new (second) UI needs to be developed. The current UI could improve too. It could be cleaner and faster, and there is space for some thoughtfully implemented JavaScript (ie. auto-submit an open record when clicking to a different one).
Remove Window Components
There are a few too many layers of objects in Waffle. Window components are not needed, all of this functionality could be rolled into the parent Window class and removing an extra layer of complexity.
Developer Scripts
Another feature to add from Rails are a couple of developers command-line scripts for creating new modules and/or applications.