TGB as a prototype tool for TGE?

#1 It depends completely and totally on your implementation of your game. There is no easy answer. If you are making a Smash-TV shooter in TGB, your code will be complicated by terrain manipulations (a 3D concern) and other issues. If you are making a platformer, your collision code will change phenomenally. The question of 2D to 3D is way too vague to help with. A million games with completely custom engines have the difficulty making the transition in the high-paid commercial world. There is little estimation that I can make except that your TorqueScript should compile fine depending on the specificity but may be horribly inaccurate in what you are trying to accomplish between the two engines. Just because something uses C doesn't mean that everything is compatable between applications. TS is the same.

#2 With a little care you can seperate enough to move core gameplay between the two. However, it generally requires that you have that in mind from the very beginning (e.g. before you write any code) for precisely the reasons David mentions. It also more then likely requires you have a good depth of knowledge of both engines, otherwise you'll just make crap decisions.

#3 @David: Yeah, that's fair enough. There are way too many different types of games to generalize about it, I suppose. I guess my main questions/concern is whether or not the TS would compile without issues and it sounds like that is probably the case.

A similiar type of game (conceptually) to the one I'm working on would be chess. So, say I was making a chess game in TGB and I ported it over to TGE. Could I assume that all of the AI and other game logic would continue to function correctly while all movement code would need to be rewritten?

@Tom: That's exactly the approach with the code that I was thinking about taking. Unfortunately though, I have a good depth of knowledge of TGB but my knowledge of TGE is nearly non-existent since I've just purschased it. Though I am pleased to see how much my conceptual understanding in TGB has helped me get acclimated to TGE. But I am worried about making those crap decisions. ;)

Maybe I'll just continue on the prototype in TGB for now and take care to organize the code well and just hope for the best when I move it over to TGE.

Thanks for the feedback guys.

#4 The only way to get used to TGE is to use it. Come up with some simple idea that touches similar code/concepts as what you actually want to do, but simplified to the point that it is achievable in a day or so per engine. Then implement that idea on each engine independantly of each other. Once you've done that for both, compare the results. From there you can build up a picture of the kind of things to look out for when you do it for real.

If you are working to any kind of schedule, I'd recommend shipping on TGB for now. If you are definately going to move over to TGE later, the sooner you learn what you need to know the quicker it will take to do the move. Continuing with the current codebase before doing the above research will only add to the probably already enormous mountain of code that you have to rewrite from scratch when you move to TGE.

#5 Depends a lot on the game too. I could imagine doing some games in TGE that work almost exactly the same, but those would be pretty simple (perhaps grid-based puzzle games).

In fact, I've got it in mind to do some work making some networked games which do exactly that. However I'm probably like you Tom and know both engines well enough to design it properly in the first place. Unless you know both, it might be a lot harder.