DotNet: Discussion on Scene Loading

Hi Folks,

As I've been chipping away at OGRE for a little while, I've been mulling over the alternative methods (that I can think of) for loading scene geometry (and the controlling logic) for multiple scenes (aka levels).

Many of the texts I've read talk about use scripts to do all kinds of things like materials (as OGRE does ), character behaviours et al.

To some extent, I can see the usefulness of this..you don't have to recompile your code each time, and I guess it could be useful for 3rd party addons if that's your gig.

But I wonder to what extent the recompile factor really plays for us DotNet folk...hell, my laptop recompiles a truckload of code in no-time-flat. So, for instance, storing level descriptions in a script doesn't appear to be a great saving *and* it doesn't prevent prying eyes/hackers from screwing with your IProperty.

So I guess what I'm asking here, is do you think it's better to use (say) a Select Case in a CreateScene sub that takes, for example an enumerated argument to construct the scene (level) that you want, and a logic sub that applies whatever logic, than bearing your sole via an editable script?

What approach(es) are you considering?

The infrastructure I've built supports both scripted and discrete models...

Cheers,

G.

From my point of view, the compile time isn't the only benefit to load various scripts. It's much simpler for a non-coder (like an artist) to write a little script. It's also easier to write an exporter for a modelling package or a custom-build tool that write to a custom file format rather than to a .cs file.

But it's all dependent of your needs: for a one-man team (or a small application), it can be quicker and fine to make it all in code instead of spending times to write a loader/exporter/custom tool/whatever.