Dev Update 4

Another 10 days! What've we got done. Well, we submitted our Imagine Cup Project Blueprint entry earlier this week, on the 17th. We think we have a fairly strong submission. 

Anyway, as promised, physics was in the works for the past 10 days. We can now run up and down stairs smoothly. It's not perfect, but it's still in the works a bit. Also, we actually started adding some of the first real content. We have a working door, and a bunch of other stuff to place - most of which doesn't actually function yet, but they look pretty! I've implemented a placeholder lighting system to pretty things up a bit. We've also set up our unit testing framework and libRocket. We're using libRocket as our GUI system.

Today, I set up an Arch Linux server, which is plugged in at our upper division computer lab in the computer science department at our university. I'll reveal some of the stuff we'll be utilizing this server for later, probably in the next update.

Oh yeah, we also finally implemented the back grid, so you can place wall tiles.

The obligatory video:

That's all for this update, thanks for reading!

More Physics, First Spaceship Concept Art

So it's been another 10 days since the last post, and I've been really stressed out. I've been constantly thinking about how huge Tetraverse is, and how much a challenge it is to implement. Its development is a huge undertaking. Anyway, we worked on the physics engine more. We got multi-manifold and grid on grid collisions working. Basically, since calculating true grid on grid collisions would be incredibly difficult to implement, we simply perform a polygon on grid collision check, using an approximated convex hull for the smaller of the two grids. Right now, the convex hull approximation isn't perfect, but it works well enough for now. Check out a video:

The new physics engine proves to be a much cleaner and smoother system than the old hard-coded collision system. Also, Ryder produced the first concept art of a very simple exploration ship. It looks pretty sweet:

For the next week, I'll be working on the physics. I'm going to fix the crack problem, where the edge of your character's physics box gets stuck on the corner of the next tile, which is visible in the video. I'll also be fixing the convex hull approximator, which will produce the smallest convex hull that contains all of the vertices. There's also a jitter bug with the polygon on grid collisions, which is being caused by the multi-manifold collisions. The same impulse is applied multiple times. The solution is to use only the smallest penetration for each axis, which you can expect in the next update.

That's all for tonight.