Chilled

Working spasmodically over the last couple of days. Felt like I needed to back off a little bit, since my mind was beginning to race constantly, and it was affecting my sleeping patterns, which in turn affected my productivity. So work performed must be measured to ensure optimal output.

Chilled out now though, thanks to good old Elliott. The relaxed mind state has assisted in persistence and patience in moments of slow progress, such as being unable to crack the nut of OpenGL texture mapping for Synergy. I've gone over every line of code a number of times, and can find no deviation from the example code I have been observing, yet my code does not work.

My only conclusion is that it is some combination of operations I am doing, which on the their own are valid, but in sequence lead to failure of texture mapping. I believe it has something to do with the way I have set up materials in combination with the parameters I have provided AllegroGL to load the bitmaps with.

After spending a couple of hours here and there each day trying to crack it, I set it aside and turned my time towards more fruitful endeavours: models and animation. Progress has been going well with these, and I have the SynMesh and SynModel classes laid out, with a little bit of body to them. I'll be sinking my teeth into the gut of them sometime soon. I have a very clear idea in my head of how the entire Skeletal Animation Method works; a depth-first traversed tree of matrices which are pushed onto a matrix stack similar to OpenGL's method of rendering.
I've also been talking with Dylan about a test bed for Synergy. That is, a game built with it to test and showcase its features. Those are key goals for the game, but we will certainly not ignore the most important goal: to make an insanely fun, insanely cool game we (and others) can play.

The concept we decided on was made unanimously, silent and in parallel with each other. We basically said at the same time `It should be Galileo Complex!'

`What's that?' you ask. Well how do I not mutate this into a huge story... basically it's a game concept we came up with a couple of years ago, that we never really had the resources and faculties to do at the time, but was always going to be a cool idea. It's a top-down action game that is a blend between death match and RTS. You control a commander who leads a group of AI-controlled players, and you must complete objectives in order to progress through the current level you are on. Objectives change depending on what other commanders have done, and matches can be a back-and-forth tug of war over areas of the level and its various objectives and resources.

I won't really go into more detail, but that's the core of it. We are inspired by the elegance of Quake's Deathmatch (Quake 1 that is), and Total Annihilation's RTS combat and interface. So our feature list pools alot of things we observe these games exhibiting; a key one being the game should be extendible. That is, people should be able to mod is easily and in a fun way. There's lots of exciting ideas we have, and I'm itching to get the Synergy Engine up to a point where we can start actually sinking our teeth into this thing.

Originally I thought the game could be an orthographic, fixed camera angle viewport. That is, essentially a 2D game. We're thinking of going with perspective rendering now, but an interesting design idea I had for the 2D thing was character-sprites that had Normal Maps, thus allowing for real-time lighting to be calculated for them, giving them a very 3D appearance without having to rasterise polygons.

While waiting for my tea to brew, I paced the living room and devised a rough algorithm for calculating the normal-mapped sprite from a 3D model.

I haven't really checked it for correctness or efficiency (should be roughly O(xy)), but it feels pretty solid. Despite the fact we're probably going with perspective projection for Galileo, I'll still implement this feature in Synergy. We'll probably end up using it in a number of places in the game anyway; it's too cool not to fiddle around with.

Now, on to Zanath. Figured out some exciting things regarding terrain creation for this. It came about when I finally got around to exporting the terrain model I've been working on to Ogre3D's .SCENE format. Basically, when you export, you get a bunch of easily-editable text files that you can do alot of tweaking with. After checking out some stuff in various forums, I believe I can get texture-splatting happening for Zanath!

This has got me excited, since it means best performance, great detail and easy asset creation/editting. I still need to do alot of experimenting and such to figure out exactly how to do it, but I feel confident that it is entirely within reach.

Until I do actually figure it out, I have to export things using megatexture-like methods. But the result is pleasingly accurate:


That's about it I guess.