Connection Success & the Client!

So this is the companion post to the Connection Success post in the main thread.

To continue from the last post, I was trying to figure out how to parse my Collada super model.  This turned out to not be as big a pain as I had previously anticipated.  It did take both Zach and myself about 3 hours to figure out how to get everything working and rendering smoothly however.  I guess I'm finding that in the world of web dev that isn't such a bad turnaround time when you are stating from scratch.

The main issue was that I had exported the super model "wrong".  I had exported the lighting and camera, etc... a.k.a. the entire scene.  Which in turn, caused the lights and everything else to be stripped out of the sub models, making them render without their skins because there was no light in "their" scene, regardless of the world light I had in the Three JS game scene.   Another way to think about it is that there are two tiers of lighting;

Tier 1: the model lighting that needs to be baked into the model (and not accidentally imported WITH it)
Tier 2: the game scene lighting that lets the model show up in the game scene.

What this looks like;

No model light, No game scene light:    Pure blackness, your CSS will display but none of your WebGL will.
Yes model light, No game scene light:   Pure blackness, your CSS will display but none of your WebGL will.
No model light, Yes game scene light:   Models will display pitch black but will be visible in the game scene.
Yes model light, Yes game scene light:  This is what success looks like. Skinned models in the game scene.

Once we got all this sorted out and realized I had imported the lights, etc... as unit models, parsing the super model became much simpler and the kinks worked out nicely.  These problems all arise from the need to understand and be able to manipulate every aspect and parameter of the model meshes crashing with my learn-as-you-go style of game creation.  This style doesn't usually cause as much headache but the learning curve to understanding the entire 3D graphics pipeline is pretty steep and a pretty focal piece of the design process.

So, as my knowledge of the 3D rendering domain grows, our path towards a playable game becomes easier to traverse.  This knowledge will eventually even guarantee a better end product.  This is because I am researching and creating every step by hand (all two man game-dev is like this) instead of using off-the-shelf models from another artist.  Which is also the reason that rigging and animation is being pushed to the end.  If we wanted to use pre-animated models we could have a much prettier prototype but it's a skill I'm not willing to pass up on learning so it's taking us a bit longer.

I think that's enough for this post.  I'll cover CORS and cross-origin permissions in the next post.

Thanks for reading,
-Clay


Oh,  here are some INSPERATIONAL pics I downloaded from the web.

I DIDN'T MAKE ANY OF THESE.

They are just what I am basing my terrain designs on (To be shown and even finished...  later... just later).