QUOTE(ChromeAngel @ Jul 24 2008, 07:14 PM)

I am more concerned with how fast LUA scripts will perform in the high performance world of twitch FPS gaming. C++ is used in source because it is so very quick, even amongst compiled languages. Replacing all that C++ with LUA worries me that it's going to slow down performance. I'm sure LUA is fine for hundreds or even thousands of lines of code, but I can't see it scaling terribly well when it comes to the hundreds of thousands of lines that it will take to implement the NS2 game code. I imagine Flayra will end up shunting some of that back into the engine for Max to optimize, namely the AI for he bots.
It's absolutely perfect for the game logic; this is just reacting to events that take place (kills, deaths, tech choices, etc.). In addition, you get the added benefit of not having to deal with null-pointer debugging. The only way Lua could slow things down is if it gets mixed into anything related to the real-time aspects of the game, which I really don't think they're going to do.
Even then, Lua is not to be underestimated; it's one of the fastest scripting languages available (
http://shootout.alioth.debian.org/gp4/benc...ll&lang=all ). I have used it to write code that runs every server frame for a Quake2 server (frames occur something like ten times a second), doing admittedly menial tasks like continually updating player positions, but I've noticed little-to-no performance impact.
You raise a good point about the client-side stuff. I don't know how much scripting support the client will get, but at the very least, I hope they design a GUI system that presents a front-end that is backed by code at the server. This would allow you to move all that !rtv, !rank, etc. sort of stuff into a nicer pop-up. This would seem to be justified just for the commander's menus, so perhaps we'll see it. I am guessing the client will be kept dumb and will get no support for reflecting on the game world via scripting.