http://blogs.msdn.com/mitchw/archive/2007/04/27/audio-compression-using-xact.aspx

But yes, I did write a separate app that allowed me to create them. Yea it was a pain, but in the end I think we were still better off than writing a complete editor from scratch just because of one lacking (and very important at that) function. The engine is a great tool, but I doubt any engine could take into account every possible use; yet, I feel the lacking functionality of creating general concave polygons is something that could be sorely missed and something many people could make use of (custom physics? animation paths?)

So now our workflow isn’t straight from TGBX to game but more like TGBX->custom poly tool->game. And no, we are not using Torque’s internal physics. We have our own (more basic) physics engine sitting on top of torqueX, which also allows us to collide against concave polys. I haven’t had much time to play with TX’s physics engine.

This was one of the things that turned me off. I wanted to create Concave polygones in the editor that I could hook into my physics engine. I didn’t want to have to write a seperate app just for that.

Did you create a seperate polygon editor? Are you using Torque’s internal physics? If so, how is it?

TorqueX has been good so far. The time saved has been invaluable. When Gunstyle is released I’ll try to write a more detailed post on my experiences with TorqueX. But in short, yes I think it’s really worth the learning curve it requires. As for a replacement for pure XNA Game Studio Express, it depends on what you are using XNA for. If it’s a full game you are trying to develop then I think TorqueX can really help you create a game closer to your original vision, and do it in far less time (learning curve included). But if it’s some quick demo app or a ‘for fun throw-away’ prototype then pure XNA GSE might be better still before transitioning to a full blown game engine.

Pros so far:
1. The editor is superb for making levels, and should suit most of your needs. (It suited all of ours except one, which was visually creating convex polygons for level geometry).
2. The component system really makes TX flexible for whatever game engine architecture fits you (in our case a networked game with components that plugged into our network layer).
3. The level of visual fidelity you get from using TX’s 2D graphics engine is much higher than what we would have been able to achieve on this tight schedule and limited shader knowledge.

Cons so far:
1. Minor tweaks could be deal breakers if you don’t have TX source and TX doesn’t expose that functionality. If we didn’t have the TX source to modify for better networked play we would’ve been out of luck it seems :\ (phew).
2. GUI system is lack-luster for PC oriented games where the mouse is the primary input. I wrote a separate, pure XNA gui library to sit on top of TX for this reason.

How do you like TorqueX now that you’ve been using it for a while? Is it worth the learning curve it requires? Would you recommend it in place of pure XNA Game Studio Express?