Suggestions for PIE (Pyro Interface Engine)

[DarkRider]

I am coding my own version of Turing's GUI library/engine. My goals for PIE (Pyro Interace Engine, my own version) are to make it fast as possible, with most, if not all, graphics handled some way by the programmer themselve. I hope to replace the majority of Turing's GUI's with my GUI's in any project I create in the future. I have found Turing's GUI's to be really buggy (and slow). Wells there are various work arounds to these bugs some bugs are hard (sometimes impossible) to handle.

Anyways what I was wondering is what you guys think would be the best way to let the user handle all graphics in PIE. I have thought of 4 ideas that I could use:

• First I could make users pass a procedure containing the graphics for each object, or group of objects, when creating the object. The problem with this is that users would have to supply a procedure for every object, or group of object, causing me to always have to pass coordinates, sizes, any text, various states and various possitions.
  
• Second I could setup the graphics procedure for each graphic I use and have the user overide that procedure with their own code or let them set the graphic procedure via another procedure. This would cause the same problems as in option one.
  
• Third I could create support for custom config files containing a list of pre-defined graphics which the user would be able to switch between via a procedure in PIE.

Currently I am making the user pass various graphic procedures into PIE, which I then call from within PIE itself. This is starting to cause me problems as I am having to pass lots of information to these procedures and the code within them are getting more and more complex. I am worried that by the time I am done coding PIE that it won't be as user-friendly as I would like it to be when creating custom graphics.

I like to hear your opinions on how to implement this kind of functionality.

[Nick]

/me votes on custom config files