Re: [Adonthell-devel] Adjustments to items.h ?

[Alexandre Courbot]

> Well, you guys are the bosses,

Hehe. The bosses of WHAT? ;) You sound like you have not your word to
say, which is wrong.

> so I'll go with whatever approach you want to 
> take, but I personally think that decentralizing game mechanics is a mistake. 
> Unless I've missed something, it looks like you want to have a series of 
> python scripts handle game mechanics functions that will be required both in 
> and out of combat.

Mmmm no, I'm not for decentralizing the game mechanism, on the contrary.
The base item class would still be a C++ class, which would allow you to
combine objects with each other, use them, etc... But when you, say, use
an object, it would refer to the Python method to know what exactly to
do. Because we want the game datas to be independant of the engine. Say
someone make a futurist RPG with our engine, he'd probably need to
combine objects with each others, but maybe not a 2 hands sword. The
objects mechanisms (or skeleton, if you prefer) are defined in C++,
while the objects themselves (their specialisation to the game their
belong to), what they actually do are rather defined in Python. This is
what I want to achieve with my idea, and I think Kai's one aims the same
goal.

So, nono, don't worry, the mechanisms still belong to the C++ side. You
can think of it as the way mapcharacters are handled in 0.3: they exist
in C++ side, there are some convenience functions (path finding, etc...)
which are coded in C++. But the actual behavior is in a Python class,
which calls these C++ functions.

Alex.

-- 
http://www.gnurou.org