Experimenting has begun with procedurally generating the terrain surface! Though still early days, I’m keen to explore the results… what’s clear is that getting something that will look superb and perform is an art form, and although I already had great respect for the work being done by others (youtube/’planetary renderer’) it’s now even more profound!
I’ll point out that this dev-diary stuff does look dull, flat, and uninteresting! That’s because there’s no point making ‘interesting’ on code and algorithms that will most likely hit the cutting room floor. This prototyping/experimentation is to help me choose the right approach for the final technique. I have a lot of requirements for my ultimate aim, and I need to be sure the approach can deliver on all fronts…
- Procedural generation
- Unique per world
- Tunable
- Over-ridable (author led content)
- Works on all platform targets
- Supports all world surfaces (barren, rock, ice, forest, craggy, desert, et al)
… to name a few 🙂
Once I’ve decided on the approach, then the interesting stuff comes in – terrain texturing, surface detail, atmospherics, flora, fauna, cities, et al. I’ll be building the worlds up ‘layer by layer’, establishing each – and then revisiting them again when the first pass is finished to refine and improve them all.
When I finally am happy, I can establish my first base on an uninhabited world. Only then shall I deem it ‘done’ 🙂
Naturally, this is a tall order. I’ll be working on this on and off throughout the dev work that still needs to be done – that way I get to work on part of the ‘end-game’ without losing sight of the fact there needs to be a ‘game’ in the first place 😉
Comments are closed.