Here’s a sneak peek at the waterfall that gives our upcoming new episode its name: Tower Fall!
https://www.youtube.com/watch?v=XZ8aA5E98p8&feature=youtu.be
We’ve been getting a lot of questions and suggestions for both Tower Fall and the game in general. Sometimes we can respond to those directly, and sometimes we’re a bit cagey. Why? Sometimes we want to save that news for some later date, but in some cases we just don’t know the answer yet. We’ve got a pretty detailed design document, but some features are lower priority and may get cut or delayed if the main features take longer than anticipated. As well, game development is really just an unending series of problem-solving, and we often don’t know exactly how we’re going to solve a problem until we’re deep into it.
For example:
Many players have suggested that we switch from four static times of day to a continuous model where time passes continuously (albeit at a faster rate than reality). In fact, nobody would like this more than me. We’ve made several games with this feature and it adds so much to the feeling of immersion.
So what’s stopping us? Technical complications. While it’s very easy to drop in a day-night cycle (as a player once told us, “I could do that in five minutes in Unity”), it’s then much harder to make everything look as good and run as well as it did before. Specifically:
- Distant views: Currently we use skyboxes which show distant scenery, such as the Lamar Valley from the ridgeline in Amethyst Mountain. These skyboxes are static renderings, one for each time of day, so we can’t use this method if the sun is always moving across the sky. Instead, we’d need actual 3D models of that surrounding environment, which is doable but obviously takes some work. It’ll also put an additional load on rendering time during gameplay and thus have a performance impact.
- Shadows: Currently we “bake” shadowmaps for each time of day. On lower quality levels, everything that doesn’t move (that is, everything except animals) uses these baked shadows. On higher quality levels, we use realtime shadows for all nearby objects and baked shadows for objects further away. This offers a great balance between visual quality and fast performance, since you still see shadows cast by distant trees and buildings, which makes the environment look so much better. But if the sun is always moving across the sky, we can’t use baked shadows — all shadows must be realtime, which means that trees and buildings more than 50-200 meters away won’t have any shadows at all (partly so the game doesn’t run too slowly, and partly for some other technical reasons).
Games often are able to minimize both issues through landscape design (e.g., limiting broad vistas in their landscapes), but alas, we are stuck with actual Yellowstone topography, which is pretty much nothing but broad vistas. Nevertheless, we are trying to solve the problem. What makes it especially hard is the fact that WolfQuest is a very cross-platform game, with people playing it on everything from powerful gaming rigs and cheap mobile phones.
We’re currently experimenting with several technical solutions, hoping that one will strike a good balance between performance and visual quality. But since we have to experiment on a lot of platforms, and because there are other related elements that are also currently in flux, it’s slow going, and so we don’t even know yet if we CAN do it satisfactorily.
tl;dr: Sometimes we don’t reply to a question because the answer is long and complicated and yet still unsatisfying. ¯\_(ツ)_/¯
