Archived Boards > FSO Overhaul Project



FAQs (to be extended as more are asked):
There're other parts of the engine that need a lot more work! We look very pretty but don't have working XYZ!
Whilst we appreciate that there are lots of parts of the engine that need work, one of the largest and most pervasive parts of the engine is the graphics and the graphics representation framework. Modularising, simplifying and improving this part of the engine will have large knock-on effects in other areas of code and hopefully allow us to simplify a lot of it.

Why not use XYZ - they already have everything we want!
Sure, but let us evaluate our needs against what they provide, how they licence their code and how difficult it would be to get it working with our codebase.

I would add a tidbit I learned from reading the coder's responses:

The Freespace SCP engine is way too interrelated. You can't recklessly modify any single part of the code since other parts could rely on the same data or use that specific implementation you're about to overwrite. (It's like house where, when you paint the living room, suddenly the carpet changes in the bathroom... and doing anything to the plumbing in the basement runs the risk of setting the kitchen on fire).

To get anywhere the engine will have to be modularized. Granted this could be done with any single "part" of the engine (like how it was done with the AI) but when you get right down to it, you'll be faced with  gargantuan monolith that casts shadows everywhere: The Rendering Engine.

So if you really want to see it another way, the projects aim is to separate this monster from the rest of the engine. (It's prime objective is not to "improve" the beast, but to pull it apart and turn it into a manageable dragon instead an all infesting hydra).


[0] Message Index

Go to full version